A logistics vehicle carbon emission optimization method, system, device and storage medium based on multi-objective collaboration

Abstract

The present invention relates to the field of carbon emission optimization technology, and in particular to a method, system, device and storage medium for optimizing carbon emissions of logistics vehicles based on multi-objective collaboration. A method for optimizing carbon emissions of logistics vehicles based on multi-objective collaboration reduces load errors and improves time synchronization accuracy through real-time correction of multi-source data; utilizes anomaly detection models to accurately identify abnormal behaviors such as overloading, high fuel consumption, and false mileage; constructs a multi-dimensional carbon efficiency feature model to effectively diagnose operational problems and reduce the incidence of overloading; uses an improved NSGA‑II algorithm to solve a multi-objective optimization model to reduce carbon emissions per ton-kilometer and increase the average daily utilization rate of vehicles; and builds a decision support mechanism to shorten solution generation time and reduce constraint violation rates. This provides logistics companies with a low-cost, quantifiable way to improve carbon efficiency in terms of environmental compliance and operational efficiency, helping them achieve green and sustainable development.

Description

Technical Field

[0001] The present invention relates to the field of carbon emission optimization technology, and in particular to a method, system, device and storage medium for optimizing carbon emissions of logistics vehicles based on multi-objective collaboration. Background Art

[0002] As the world actively responds to climate change, the logistics industry, as a major carbon emitter, faces a critical challenge in carbon efficiency management. Currently, logistics companies face numerous challenges in carbon emissions monitoring and management. For one thing, data fragmentation is a prominent issue. Logistics operations involve multiple links, including transportation, warehousing, and distribution. Data from each link is stored in separate systems, such as weighing data from truck scales, access control timestamps, fuel consumption data, and mileage data. The lack of an effective integration mechanism makes it difficult for companies to obtain comprehensive and accurate data to support carbon emissions accounting and analysis, resulting in poor carbon accounting accuracy. Furthermore, existing carbon accounting methods rely on manual data collection and simple calculations, which are susceptible to data errors and human factors, and cannot accurately reflect the true carbon emissions of logistics activities. These issues not only hinder logistics companies from effectively controlling their own carbon emissions, but also limit their progress in environmental compliance, making it difficult to meet increasingly stringent environmental regulations.

[0003] Traditional decision-making mechanisms are no longer adaptable to the needs of business development. In logistics operations decisions, companies often focus solely on single objectives such as transportation costs and efficiency, overlooking the synergistic relationship between carbon emissions and operational efficiency. This results in incomplete and unscientific decision-making. For example, in vehicle scheduling and task allocation, the impact of vehicle load and speed on carbon emissions and energy efficiency is not fully considered. This leads to frequent instances of vehicle overloading and inefficient driving, which not only increases carbon emissions but also reduces the company's overall operational efficiency. Furthermore, due to the lack of effective anomaly detection mechanisms, logistics companies struggle to detect and address abnormal behaviors such as overloading, high fuel consumption, and false mileage in real time, further exacerbating resource waste and increased carbon emissions. This outdated decision-making mechanism makes it difficult for logistics companies to strike a balance between environmental protection and economic development. An innovative system and method is urgently needed to optimize decision-making and achieve the synergistic development of multiple objectives, including reducing carbon emissions and improving operational efficiency. Summary of the Invention

[0004] To solve the above technical problems, the present invention provides a logistics vehicle carbon emission optimization method based on multi-objective collaboration, which is characterized by comprising the following steps:

[0005] S1. Obtain a multi-source data set of logistics vehicles, a set of evaluation indicators of logistics vehicles, and a data set of logistics vehicle tasks, and perform data correction on the multi-source data set of logistics vehicles through a Kalman filter model;

[0006] S2. Based on the corrected logistics vehicle multi-source dataset, a joint probability density model is constructed to perform anomaly detection on the corrected logistics vehicle multi-source dataset to obtain logistics vehicle anomaly data;

[0007] S3. Based on the abnormal data of logistics vehicles and the set of logistics vehicle evaluation indicators, a multi-dimensional carbon efficiency characteristic model is constructed to obtain a carbon efficiency characteristic matrix;

[0008] S4. Based on the carbon efficiency characteristic matrix, a multi-objective carbon efficiency optimization model is constructed, and the multi-objective carbon efficiency optimization model is solved using the improved NSGA-II algorithm to obtain a Pareto frontier solution set;

[0009] S5. Based on the logistics vehicle task dataset, a dynamic weight scoring model is constructed to evaluate the Pareto frontier solution set, obtain the optimal Pareto solution set, and implement it as a logistics vehicle carbon emission optimization strategy.

[0010] Furthermore, the multi-source data set of logistics vehicles includes vehicle load data, vehicle access control system data, vehicle fuel consumption data, vehicle transportation mileage data and vehicle basic operation data;

[0011] The logistics vehicle evaluation index set is of the same data type as the logistics vehicle multi-source dataset, and is used to perform data evaluation on the logistics vehicle multi-source dataset;

[0012] The logistics vehicle task data set includes task allocation data and business constraint data;

[0013] The basic vehicle operation data includes vehicle type, maximum vehicle load, vehicle fuel type and vehicle carbon emission coefficient;

[0014] The business constraint data includes business priority, real-time road condition data and vehicle maintenance data.

[0015] Furthermore, the step S2 includes the following steps:

[0016] S201. Based on the corrected logistics vehicle multi-source data set, a logistics vehicle multi-source data matrix is ​​constructed, and a kernel density estimation algorithm is used to construct a joint probability density function;

[0017] S202. Solve the joint probability density function by the local outlier factor algorithm to calculate the local outlier factor of each data point in the multi-source data matrix of logistics vehicles;

[0018] S203. Based on the obtained local outlier factor, the data points are classified as abnormal using the K-means clustering algorithm to obtain abnormal data of logistics vehicles.

[0019] Furthermore, the local outlier factor algorithm uses Manhattan distance to measure spatial distance.

[0020] Furthermore, step S3 includes the following steps:

[0021] S301. Based on the abnormal data of logistics vehicles and the set of logistics vehicle evaluation indicators, a multi-dimensional carbon efficiency characteristic model is constructed;

[0022] S302. Based on the multi-dimensional carbon efficiency feature model, perform feature association analysis using the DBSCAN clustering algorithm to obtain a carbon efficiency feature matrix.

[0023] The present invention also provides a logistics vehicle carbon emission optimization system based on multi-objective collaboration, which is implemented based on any of the logistics vehicle carbon emission optimization methods based on multi-objective collaboration described above, and includes a multi-source data acquisition module, an anomaly detection module, a multi-dimensional carbon efficiency feature extraction module, a multi-objective dynamic optimization module, and a dynamic decision support module that are sequentially signal-connected;

[0024] The multi-source data acquisition module is used to obtain a multi-source data set of logistics vehicles, a logistics vehicle evaluation index set and a logistics vehicle task data set, and perform data correction on the multi-source data set of logistics vehicles through a Kalman filter model;

[0025] The anomaly detection module is used to construct an anomaly detection model based on the corrected logistics vehicle multi-source data set through the kernel density estimation algorithm and the local outlier factor algorithm, perform anomaly detection on the corrected logistics vehicle multi-source data set, and obtain logistics vehicle anomaly data;

[0026] The multi-dimensional carbon efficiency feature extraction module is used to construct a multi-dimensional carbon efficiency feature model based on the abnormal data of logistics vehicles and the logistics vehicle evaluation index set to obtain a structured carbon efficiency feature matrix;

[0027] The multi-objective dynamic optimization module is used to construct a multi-objective carbon efficiency optimization model based on the structured carbon efficiency feature matrix, and solve the multi-objective carbon efficiency optimization model using the improved NSGA-II algorithm to obtain a Pareto frontier solution set;

[0028] The dynamic decision support module is used to construct a dynamic weight scoring model based on the logistics vehicle task data set to evaluate the Pareto frontier solution set, obtain the optimal Pareto solution set, and implement it as a logistics vehicle carbon emission optimization strategy.

[0029] Furthermore, the system also includes a visualization module for real-time display of logistics vehicle multi-source data sets, logistics vehicle evaluation index sets and logistics vehicle task data sets, and real-time display of carbon emission detection results and optimization plans.

[0030] The present invention also provides a computer device, comprising a memory, a processor, and a computer program stored in the memory and runnable on the processor, characterized in that when the processor executes the computer program, the logistics vehicle carbon emission optimization method based on multi-objective collaboration as described in any of the above items is implemented.

[0031] The present invention also provides a storage medium on which a computer program is stored, characterized in that when the computer program is executed by a processor, it implements the logistics vehicle carbon emission optimization method based on multi-objective collaboration as described in any of the above items.

[0032] The beneficial effects of the present invention are:

[0033] Through real-time correction of multi-source data, load errors are reduced and time synchronization accuracy is improved. Anomaly detection models are used to accurately identify abnormal behaviors such as overloading, high fuel consumption, and false mileage. A multi-dimensional carbon efficiency feature model is constructed to effectively diagnose operational issues and reduce the incidence of overloading. An improved NSGA-II algorithm is used to solve multi-objective optimization models, reducing carbon emissions per ton-kilometer and increasing the average daily vehicle utilization rate. A decision support mechanism is established to shorten solution generation time and reduce constraint violation rates. This provides logistics companies with a low-cost, quantifiable approach to improving carbon efficiency in terms of environmental compliance and operational efficiency, helping them achieve green and sustainable development. BRIEF DESCRIPTION OF THE DRAWINGS

[0034] Figure 1 , a flow chart of a logistics vehicle carbon emission optimization method based on multi-objective collaboration of the present invention.

[0035] Figure 2 , a structural schematic diagram of a logistics vehicle carbon emission optimization system based on multi-objective collaboration of the present invention.

[0036] Figure 3 , a scatter plot of carbon efficiency of an embodiment of the present invention.

[0037] Figure 4 , a Pareto front solution set distribution diagram of an embodiment of the present invention.

[0038] Figure 5 , An embodiment of the present invention proposes a schematic diagram of the terminal equipment structure of a logistics vehicle carbon emission optimization method based on multi-objective collaboration.

[0039] Figure 6 , a schematic diagram of the computer-readable storage medium structure of the logistics vehicle carbon emission optimization method based on multi-objective collaboration proposed in an embodiment of the present invention.

[0040] In the figure, 200 - terminal device, 210 - memory, 211 - RAM, 212 - cache memory, 213 - ROM, 214 - program / utility, 215 - program module, 220 - processor, 230 - bus, 240 - external device, 250 - I / O interface, 260 - network adapter, 300 - program product. DETAILED DESCRIPTION

[0041] In order to enable those skilled in the art to better understand the contents of the present invention and make the objectives, technical solutions and advantages of the present invention clearer, the present invention is further described in detail below in conjunction with embodiments and drawings.

[0042] The exemplary embodiments of the present invention and their description are only used to explain the present invention and are not intended to further limit the present invention.

[0043] Example 1:

[0044] like Figure 1 As shown, an embodiment of the present invention provides a method for optimizing carbon emissions of logistics vehicles based on multi-objective collaboration, which is characterized by comprising the following steps:

[0045] S1. Obtain a multi-source data set of logistics vehicles, a set of evaluation indicators of logistics vehicles, and a data set of logistics vehicle tasks, and perform data correction on the multi-source data set of logistics vehicles through a Kalman filter model;

[0046] S2. Based on the corrected logistics vehicle multi-source dataset, a joint probability density model is constructed to perform anomaly detection on the corrected logistics vehicle multi-source dataset to obtain logistics vehicle anomaly data;

[0047] S3. Based on the abnormal data of logistics vehicles and the set of logistics vehicle evaluation indicators, a multi-dimensional carbon efficiency characteristic model is constructed to obtain a carbon efficiency characteristic matrix;

[0048] S4. Based on the carbon efficiency characteristic matrix, a multi-objective carbon efficiency optimization model is constructed, and the multi-objective carbon efficiency optimization model is solved using the improved NSGA-II algorithm to obtain a Pareto frontier solution set;

[0049] S5. Based on the logistics vehicle task dataset, a dynamic weight scoring model is constructed to evaluate the Pareto frontier solution set, obtain the optimal Pareto solution set, and implement it as a logistics vehicle carbon emission optimization strategy.

[0050] Furthermore, the multi-source data set of logistics vehicles includes vehicle load data, vehicle access control system data, vehicle fuel consumption data, vehicle transportation mileage data and vehicle basic operation data;

[0051] The logistics vehicle evaluation index set is of the same data type as the logistics vehicle multi-source dataset, and is used to perform data evaluation on the logistics vehicle multi-source dataset;

[0052] The logistics vehicle task data set includes task allocation data and business constraint data;

[0053] The basic vehicle operation data includes vehicle type, maximum vehicle load, vehicle fuel type and vehicle carbon emission coefficient;

[0054] The business constraint data includes business priority, real-time road condition data and vehicle maintenance data.

[0055] Furthermore, step S1 includes the following sub-steps:

[0056] S101. Construct a state vector, expressed as:

[0057]

[0058] in, is the state vector at time k, is the actual cargo weight at time k, is the weight change rate at time k;

[0059] S102. Construct the observation equation, expressed as:

[0060]

[0061] in, is the observation value at time k, H is the observation matrix, is the observation noise, and the observation noise It obeys a normal distribution with a mean of 0 and an observation noise variance of R;

[0062] S103. Define the system dynamics equation, expressed as:

[0063]

[0064] Where F is the state transfer matrix, is the state vector at the previous moment, and , time interval Δt=0.2s; , represents the system process noise, which obeys the normal distribution with mean 0 and process noise covariance matrix Q. The process noise covariance matrix .

[0065] Specifically, the implementation principle flow of each sub-step in the above embodiment is as follows:

[0066] First, we built a data collection system, including scale data collection: connecting to the Keli D2008 weighing system via the RS485 interface, setting a high-frequency sampling frequency of 1Hz to obtain raw cargo weight data (unit: kg); access control data collection: reading the vehicle entry and exit timestamps (format: YYYY-MM-DD HH:MM:SS.sss) from the Hikvision access control system through the API interface, with an accuracy of ±0.1 second.

[0067] Then, construct the Kalman filter model. State variables: Definition ,in, is the state vector at time k, represents the actual cargo weight at time k, represents the weight change rate at time k (kg / s), with a dynamic range of ±50kg / s; the observation equation is: ,in, is the observation value at time k; is the observation matrix, which is used to map the state vector to the observation space; represents the observation noise, and , that is, the observation noise obeys a normal distribution with a mean of 0 and a variance of R; the observation noise variance (FS = 10 tons), where FS stands for full mileage, which is used to indicate the maximum value that the measuring instrument can measure within the measuring range; system dynamics model: , where the state transfer matrix ,in = 0.2s, indicating the time interval between two adjacent moments. State transition matrix Used to predict the current state based on the state of the previous moment; process noise covariance matrix , used to quantify the impact of process noise on the system state;

[0068] Next, perform filtering calculation and prediction: ,in, is the estimated state value at time k, is the state vector at the previous moment, , in, is the error covariance matrix of the previous moment, represents the error covariance matrix of the current moment of prediction, is the transposed matrix of the state transfer matrix; update step: , ;

[0069] Then, parameter optimization is performed and the optimal parameters are determined through sensitivity analysis: process noise covariance matrix (Influence weight is 0.78); Observation noise variance R = 22500 (Influence weight is 0.22); Δt = 0.2 seconds;

[0070] Finally, we conducted a validation test on 30 trucks (10 each of 8-ton, 15-ton, and 30-ton capacity) for seven consecutive days, comparing the following indicators:

[0071] Load error rate: 15%-20% before correction → ±0.8% after correction;

[0072] Time synchronization accuracy: ±5 seconds before correction → ±1 second after correction;

[0073] The error calculation can be expressed as: , where WCZ is the error value, For the measured load, The actual load, time synchronization error: the absolute value of the difference between the access control timestamp and the weighbridge timestamp. The verification data is shown in Table 1:

[0074] Table 1 Error results table

[0075]

[0076] Furthermore, the step S2 includes the following steps:

[0077] S201. Based on the corrected logistics vehicle multi-source data set, a logistics vehicle multi-source data matrix is ​​constructed, and a kernel density estimation algorithm is used to construct a joint probability density function. The joint probability density function is expressed as:

[0078]

[0079] in, is the kernel density, which represents the multi-source data of logistics vehicles, W is the vehicle load, w is the vehicle load at the estimated joint probability density, L is the vehicle fuel consumption, l is the vehicle fuel consumption at the estimated joint probability density, S is the vehicle mileage, s is the vehicle mileage at the estimated joint probability density, h W is the bandwidth parameter corresponding to the vehicle load data dimension, h L is the bandwidth parameter corresponding to the vehicle fuel consumption data dimension, h S is the bandwidth parameter corresponding to the vehicle mileage data dimension, K is the kernel function, n is the number of logistics vehicles, i represents the i-th sample data, W i is the vehicle load of the i-th sample data, L i is the vehicle fuel consumption of the i-th sample data, S i is the vehicle mileage of the i-th sample data;

[0080] S202. Solve the joint probability density function using the local outlier factor algorithm to calculate the local outlier factor of each data point in the multi-source data matrix of logistics vehicles. The local outlier factor algorithm is expressed as:

[0081]

[0082]

[0083] in, is the i-th sample data, for The local outlier factor value, o is The number of neighboring points of for The jth neighbor of For distance The set of the most recent o data points, for The local reachable density of for The local reachable density of

[0084] S203. Based on the obtained local outlier factor, the data points are classified as abnormal using the K-means clustering algorithm to obtain abnormal data of logistics vehicles.

[0085] Specifically, the implementation principle flow of each sub-step in the above embodiment is as follows:

[0086] First, logistics vehicle data integration and preprocessing are carried out, including:

[0087] 1) Vehicle load (W): obtained based on filtered data from the scale, with a data error of ±0.8%;

[0088] 2) Vehicle fuel consumption (L): The initial value is reported by the logistics vehicle driver. The reported value is verified using the CRC algorithm and the incorrect fuel consumption data is eliminated. The unit is L / 100km.

[0089] 3) Vehicle mileage (S): obtained by matching the access control system with the transportation documents, with a data error of ±3%;

[0090] 4) Theoretical fuel consumption calculation: obtained through historical data regression, which can be expressed as:

[0091]

[0092] Among them L ll is the theoretical fuel consumption, a and b are the obtained regression coefficients, and c is the regression error;

[0093] Then, the joint probability density modeling is performed through the kernel density estimation theory KDE. Specifically, the KDE parameters are set as follows:

[0094] 1) Sample size: Using 90 days of normal operation data, the data volume of logistics vehicle trips n is 2700;

[0095] 2) Kernel function: Using Gaussian kernel function, it can be expressed as:

[0096] ;

[0097] Among them, K(u) is the Gaussian kernel function, e is the natural logarithm, and u is the standardized distance;

[0098] 3) Bandwidth optimization: According to Silverman's bandwidth rule, = 50kg, = 1.2L / 100km, = 8km;

[0099] 4) Construct joint probability density function:

[0100]

[0101] in, is the kernel density, W is the vehicle load, w is the vehicle load at the estimated joint probability density, L is the vehicle fuel consumption, l is the vehicle fuel consumption at the estimated joint probability density, S is the vehicle mileage, s is the vehicle mileage at the estimated joint probability density, h W is the bandwidth parameter corresponding to the vehicle load data dimension, h L is the bandwidth parameter corresponding to the vehicle fuel consumption data dimension, h S is the bandwidth parameter corresponding to the vehicle mileage data dimension, K is the kernel function, n is the number of logistics vehicles, i represents the i-th sample data, W i is the vehicle load of the i-th sample data, L i is the vehicle fuel consumption of the i-th sample data, S i is the vehicle mileage of the i-th sample data;

[0102] Next, the local outlier factor is calculated.

[0103] Parameter setting: the number of neighboring points o=15 neighbors, and the distance metric uses Manhattan distance;

[0104] Specifically, the calculation steps are:

[0105] Calculate o-distance: ; Equal to the 15th smallest two-norm distance, the Represents sample data The distance to its 15th nearest neighbor is used to measure the spatial distance between data points. is the two-norm distance.

[0106] Calculate the reachable distance reach-distk, the calculation formula is:

[0107]

[0108] in, for and The accessible distance between for o-distance, taking into account the data points The k-distance and and The actual distance between them can be calculated to avoid misjudgment due to local density differences.

[0109] Calculate local reachability density

[0110]

[0111] in, for The local reachable density reflects the sample data The density of the surrounding area. The higher the local reachable density, the denser the data around the point.

[0112] Calculate the LOF value:

[0113] To judge the data points Is it an outlier? If the LOF value is much greater than 1, it means that the density around the point is significantly different from that of the neighboring points, and it may be an outlier.

[0114] After that, abnormality determination and classification are performed.

[0115] Abnormal level threshold:

[0116]

[0117] YC is the abnormality level. Specifically, when LOF>3.0, the abnormality level is severe abnormality; when 2.5<LOF≤3.0, the abnormality level is moderate abnormality; when LOF≤2.5, the abnormality level is normal.

[0118] According to the obtained abnormal level, the abnormal type is identified, specifically:

[0119] The identification conditions for overload and high fuel consumption are: W>1.1Wmax, L>25% of the theoretical value, where W is the vehicle load, Wmax is the maximum vehicle load, and L is the vehicle fuel consumption; the identification conditions for false mileage declaration are: S is inconsistent with the access control time, and L / S is abnormally low, where S is the vehicle mileage and L / S is the ratio of vehicle fuel consumption to vehicle mileage; the identification conditions for inefficient path are: W is normal and ,in, is the average fuel consumption per kilometer, is the standard deviation of fuel consumption per kilometer; the abnormal identification condition for idling fuel consumption is: S<50km, but L>30L / 100km.

[0120] The abnormal behavior recognition results are shown in Table 2:

[0121] Table 2 Abnormal type identification table

[0122]

[0123] Theoretical fuel consumption calculation: Based on vehicle parameters and historical data regression (e.g. 15-ton truck: ,in Indicates theoretical fuel consumption). Abnormality type distribution: severe abnormality accounts for 18%, moderate abnormality accounts for 24%, and normal accounts for 58%.

[0124] Step 3: Multi-dimensional carbon efficiency characteristic modeling and diagnosis

[0125] First, the data is integrated and preprocessed. The data sources are:

[0126] Basic operational data, including vehicle type (8T / 15T / 30T) and license plate number; dynamic transportation data, including vehicle load W, vehicle fuel consumption L, vehicle mileage S, and vehicle driving time T; environmental parameters, including the fuel carbon emission coefficient P. In this embodiment, the fuel carbon emission coefficient corresponding to diesel is used, P=2.68; manually annotated data, including historical anomaly records (such as overloading and route deviation).

[0127] Secondly, calculate the theoretical fuel consumption of current logistics vehicles. The calculation formula is:

[0128]

[0129] in, is the theoretical fuel consumption of the vehicle, W is the vehicle load, and S is the vehicle mileage;

[0130] Next, calculate the carbon efficiency index of current logistics vehicles and obtain Figure 3 The carbon efficiency scatter plot shown, specifically, includes the following data calculations:

[0131] Carbon emission rate (kg-CO2 / h): ,in is the carbon emission efficiency, P is the fuel carbon emission coefficient, L is the vehicle fuel consumption, and T is the vehicle driving time;

[0132] Energy efficiency index (t・km / L): ,in, is the energy efficiency index;

[0133] Load balance: ,in The load balance reflects the uniformity of load distribution in the transportation task. is the average vehicle load, is the standard deviation of vehicle load W;

[0134] Fuel consumption - load elastic coefficient: ,in, is the fuel consumption-load elastic coefficient, For standard load;

[0135] Time deviation: ,in is the time deviation, This is the standard shipping time;

[0136] Based on the above calculated values, DBSCAN cluster analysis is performed on the current logistics vehicles, which specifically includes the following steps:

[0137] First, set the parameters, specifically, including: standardizing the neighborhood radius to obtain = 0.25 (after normalization), where is the neighborhood radius; the minimum number of samples minPts = 5; the input features are normalized (Z-score): (Z-score normalization).

[0138] Then perform cluster analysis, including: Data standardization: ,in is the standardized input feature (including ), For unstandardized input features (including ), is the mean of the input features, is the standard deviation of the input feature; density accessibility judgment: core point → boundary point → noise point;

[0139] According to the cluster analysis results, the abnormal pattern of the current logistics vehicles is identified as follows:

[0140] Overload and high consumption type: and ,in is the mean of the input features, is the standard deviation of the input features, is the fuel consumption-load elastic coefficient;

[0141] Path inefficiency: and ,in is the lower quartile of EEI, that is, the value at the 25% position, is the interquartile range of EEI (the difference between the upper and lower quartiles);

[0142] Driving behavior type: and ,in is the carbon emission rate fluctuation value;

[0143] The distribution of abnormal types was as follows: overloaded and high-consumption types accounted for 24%, inefficient paths accounted for 18%, and normal types accounted for 58%; verification indicators: model precision rate was 91%, and the overload incidence rate dropped from 18.3% to 5.1%.

[0144] Step 4: MCMC simulation and contribution calculation

[0145] 1. Data integration and parameter setting

[0146] Input data: Vehicle attributes: Maximum load of 52 trucks (8T / 15T / 30T) , carbon emission coefficient (kg-CO2 / L), fuel consumption benchmark (L / 100km). i is the i-th sample data, which corresponds to the vehicle number in Table 2 above in step 4 of this embodiment;

[0147] Transport tasks: 150 tasks per day, including cargo weight requirements ,mileage , time window ,in For the shortest vehicle transport time, is the longest vehicle transportation time; j represents the jth task

[0148] Abnormal marking: vehicle classification (red / orange / yellow), corresponding marking weight for each classification is (red=0.5, orange=0.3, yellow=0.1).

[0149] Key parameter: load penalty coefficient ; Time window relaxation ; Mark weight .

[0150] 2. Mathematical model construction

[0151] Decision variables:

[0152] , whether vehicle i performs task c;

[0153] , the average speed of vehicle i performing task c.

[0154] Objective function:

[0155] Total carbon emissions are expressed as:

[0156]

[0157] in For the total carbon emissions target, is the vehicle fuel consumption of sample data i under the current target, is the vehicle load of sample data i under the current target, is the vehicle mileage of sample data i under the current target, It is a 01 variable that indicates whether the sample data i executes task c under the current target;

[0158] The overall efficiency is expressed as:

[0159]

[0160] in, For overall efficiency, is the mark weight;

[0161] Constraints include: load constraints ; Time window: ,in is the transportation time of sample data i under the current target; the task covers: ; Vehicle Capabilities: .

[0162] 3. Improved NSGA-II algorithm:

[0163] Population initialization: The proportion of marked vehicles in the initial solution increased by 20%;

[0164] Crossover operation: The retention rate of the red-marked vehicle task sequence is 80%; Mutation strategy: Non-uniform mutation of the speed variable (the step size decays with the number of iterations).

[0165] Optimization strategy:

[0166] The number of iterations is 200 and the population size is 300; non-dominated sorting and crowding calculation ensure the diversity of the solution set.

[0167] 4. Decision-making plan generation

[0168] Pareto frontier solution set: contains 47 non-dominated solutions, with carbon emissions ranging from 26,200 to 30,500 kg and comprehensive efficiency ranging from 76.5 to 98.3 t·km / L·h;

[0169] 5. Parameter sensitivity analysis:

[0170] = 0.5, the carbon emissions of red vehicles decreased by 22.4%;

[0171] =0.3, the task completion rate is 97% and carbon emissions increase by 2.1%.

[0172] 6. Optimization effect: Red-marked vehicles saw the largest reduction in carbon emissions (16%-24%), and overall efficiency increased by 17%-21%. Comparative indicators: Carbon emissions per ton-kilometer decreased from 0.148kg to 0.122kg, and the average daily vehicle utilization rate increased from 68% to 79%.

[0173] The verification data is shown in Table 3

[0174] Table 3 Schematic table of verification data

[0175]

[0176] Step 5: Pareto frontier solution set generation data integration multi-objective optimization model output of 300 non-dominated solutions, including: total carbon emission target ; Overall efficiency Constraint violation degree , where CV is the constraint violation, CZL is the overload, and TWPL is the time window deviation.

[0177] Some sample data are shown in Table 4.

[0178] Table 4 Pareto frontier solutions

[0179]

[0180] Pareto front generation

[0181] Non-dominated sorting:

[0182] (1) Define the dominance relationship: If and only if and , and at least one strict inequality holds, where X is Any 01 matrix, 、 Respectively Two arbitrary 01 matrices, 、 Indicates that two arbitrary 01 matrices are brought into Total carbon emission objective function and The comprehensive efficiency function is compared;

[0183] (2) Classification of frontier levels: : A solution that is not dominated by any solution; :Only The solution that governs ; and so on.

[0184] Congestion calculation: For each target m, the solution is Arrange in ascending order; the boundary decongestion degree is set to infinity; the intermediate decongestion degree can be expressed as: ;in, is the intermediate solution congestion, m is the number of targets, and there are only two targets here, one is total carbon emissions and the other is comprehensive efficiency. is the total carbon emission or comprehensive efficiency function, is the maximum value of total carbon emissions or comprehensive efficiency function is the minimum value of the total carbon emissions or comprehensive efficiency function; is the i-th sample data, is the i+1th sample data, is the i-1th sample data.

[0185] Solution set characteristic analysis:

[0186] The characteristic analysis of the Pareto front solution is as follows: Figure 4 The Pareto frontier solution set distribution is shown in Table 5.

[0187] Table 5. Pareto frontier solution set distribution statistics

[0188]

[0189] Typical solution classification: Radical emission reduction (sacrifice 12% efficiency for 19% carbon emission reduction): ( <26,500); Balanced (carbon reduction 14% + efficiency increase 11%): (27,000< <28,000); High-efficiency priority (efficiency increased by 22%, carbon emissions only reduced by 5%): ( >95).

[0190] Verification and optimization

[0191] The convergence verification results are shown in Table 6:

[0192] Table 6 Convergence acceptance results

[0193]

[0194] The optimization results of marked vehicles are shown in Table 7:

[0195] Table 7 Marked vehicle optimization results

[0196]

[0197] Solution type distribution: radical emission reduction type accounts for 24%, balanced type accounts for 48%, and high efficiency priority type accounts for 28%; constraint violation degree: when the population size is 300, 94% of the solutions have CV=0.

[0198] Step 6: Implementation of decision support mechanism

[0199] 1. Data integration and parameter setting; core data include: Pareto solution set: 47 non-dominated solutions, including carbon emissions (kg-CO2), comprehensive efficiency (t・km / L・h) and vehicle-task allocation scheme ; , whether vehicle i performs task c; arbitrary 01 matrix business constraints: customer priority (VIP / ordinary), real-time traffic congestion index , vehicle maintenance plan.

[0200] Decision parameters include: Carbon emission weight , efficiency weight ; Urgency coefficient .

[0201] 2. Decision rule construction; solution scoring model:

[0202] in, is the score of the g-th solution, is the constraint violation degree of the g-th solution. The model comprehensively considers carbon emissions, efficiency, and constraint violations to provide a quantitative basis for decision-making.

[0203] 3. Dynamically adjust rules;

[0204] According to different business needs, formulate dynamic adjustment rules: when the urgency coefficient ( =1.8), the rule applies the time-sensitive mode: when the task time is tight, priority is given to ensuring transportation efficiency, and the weight is adjusted to ; When the urgency coefficient ( =1.0), the rule applies the emission reduction priority mode: when facing environmental pressure or enterprises focus on emission reduction, the weight is adjusted to , giving priority to reducing carbon emissions;

[0205] Special rules for marking vehicles are shown in Table 8:

[0206] Table 8 Special rules for marking vehicles

[0207]

[0208] 4. Decision-making process; initial screening: elimination Solutions and vehicle maintenance plans;

[0209] 5. Scenario matching: According to different business scenarios, select the appropriate solution range, as shown in Table 9, where Remaining time for the task:

[0210] Table 9 Business scenario scope

[0211]

[0212] 6. Final decision: Determine the final implementation plan by comprehensively considering the plan scores and driver satisfaction:

[0213] ;in, is the index of the optimal solution, For all The set of g with the maximum value is obtained, and MY is the driver satisfaction;

[0214] 7. Effectiveness Verification; Decision-making Efficiency: Single-plan generation time was reduced from 45 minutes to 2.3 minutes, and the number of constraint violations was reduced from 5.2 times / day to 0.7 times / day. The target achievement rate is shown in Table 10:

[0215] Table 10 Target achievement rate

[0216]

[0217] The verification data is shown in Table 11:

[0218] Table 11 Decision implementation verification data

[0219]

[0220] Decision results: 97% of solutions were generated within 2.3 minutes, with a constraint violation rate of only 3%; driver satisfaction: modeled based on historical task completion quality, with a value range of 0-1.

[0221] Example 2

[0222] As a preferred implementation of the above embodiment, Figure 2As shown, a logistics vehicle carbon emission optimization system based on multi-objective collaboration is provided. The system is implemented based on any one of the logistics vehicle carbon emission optimization methods based on multi-objective collaboration described above, and includes a multi-source data acquisition module, an anomaly detection module, a multi-dimensional carbon efficiency feature extraction module, a multi-objective carbon efficiency optimization module, and a dynamic decision support module that are sequentially connected by signals;

[0223] The multi-source data acquisition module is used to obtain a multi-source data set of logistics vehicles, a logistics vehicle evaluation index set and a logistics vehicle task data set, and perform data correction on the multi-source data set of logistics vehicles through a Kalman filter model;

[0224] The anomaly detection module is used to construct an anomaly detection model based on the corrected logistics vehicle multi-source data set through the kernel density estimation algorithm and the local outlier factor algorithm, perform anomaly detection on the corrected logistics vehicle multi-source data set, and obtain logistics vehicle anomaly data;

[0225] The multi-dimensional carbon efficiency feature extraction module is used to construct a multi-dimensional carbon efficiency feature model based on the abnormal data of logistics vehicles and the logistics vehicle evaluation index set to obtain a structured carbon efficiency feature matrix;

[0226] The multi-objective carbon efficiency optimization module is used to construct a multi-objective carbon efficiency optimization model based on the structured carbon efficiency feature matrix, and solve the multi-objective carbon efficiency optimization model using the improved NSGA-II algorithm to obtain a Pareto frontier solution set;

[0227] The dynamic decision support module is used to construct a dynamic weight scoring model based on the logistics vehicle task data set to evaluate the Pareto frontier solution set, obtain the optimal Pareto solution set, and implement it as a logistics vehicle carbon emission optimization strategy.

[0228] Furthermore, the system also includes a visualization module for real-time display of logistics vehicle multi-source data sets, logistics vehicle evaluation index sets and logistics vehicle task data sets, and real-time display of carbon emission detection results and optimization plans.

[0229] Example 3

[0230] like Figure 5 Based on Example 1, this example proposes a terminal device for optimizing carbon emissions of logistics vehicles based on multi-objective collaboration. The terminal device 200 includes at least one memory 210, at least one processor 220, and a bus 230 connecting different platform systems.

[0231] The memory 210 may include a readable medium in the form of a volatile memory, such as a RAM 211 and / or a cache memory 212 , and may further include a ROM 213 .

[0232] Among them, the memory 210 also stores a computer program, which can be executed by the processor 220, so that the processor 220 executes any of the above-mentioned methods for optimizing carbon emissions of logistics vehicles based on multi-objective collaboration in the embodiments of this application. Its specific implementation method is consistent with the implementation method and the technical effect achieved in the embodiments of the above-mentioned method, and some of the contents are not repeated here. The memory 210 can also include a program / utility 214 having a set (at least one) of program modules 215. Such program modules include but are not limited to: an operating system, one or more application programs, other program modules, and program data. Each of these examples or some combination may include the implementation of a network environment.

[0233] Accordingly, the processor 220 may execute the aforementioned computer programs, as well as the program / utility 214 .

[0234] The bus 230 may represent one or more of several types of bus structures, including a memory bus or memory controller, a peripheral bus, an accelerated graphics port, a processor, or a local bus using any of a variety of bus architectures.

[0235] The terminal device 200 can also communicate with one or more external devices 240, such as keyboards, pointing devices, Bluetooth devices, etc., and can also communicate with one or more devices that can interact with the terminal device 200, and / or communicate with any device that enables the terminal device 200 to communicate with one or more other computing devices (such as routers, modems, etc.). Such communication can be carried out through the I / O interface 250. In addition, the terminal device 200 can also communicate with one or more networks (such as local area networks (LANs), wide area networks (WANs) and / or public networks, such as the Internet) through the network adapter 260. The network adapter 260 can communicate with other modules of the terminal device 200 through the bus 230. It should be understood that although not shown in the figure, other hardware and / or software modules can be used in conjunction with the terminal device 200, including but not limited to: microcode, device drivers, redundant processors, external disk drive arrays, RAID systems, tape drives, and data backup storage platforms.

[0236] Example 4

[0237] like Figure 6 Building on Example 1, this example proposes a computer-readable storage medium for a method for optimizing carbon emissions of logistics vehicles based on multi-objective collaboration. The computer-readable storage medium stores instructions that, when executed by a processor, implement any of the aforementioned methods for optimizing carbon emissions of logistics vehicles based on multi-objective collaboration. The specific implementation methods and technical effects achieved are consistent with those described in the aforementioned method examples, and some details are omitted here.

[0238] Figure 3 The program product 300 provided in this embodiment for implementing the above method is shown. It can use a portable compact disc read-only memory (CD-ROM) and include program code, and can be run on a terminal device, such as a personal computer. However, the program product 300 of the present invention is not limited to this. In this embodiment, the readable storage medium can be any tangible medium that contains or stores a program, and the program can be used by or in conjunction with an instruction execution system, device, or device. The program product 300 can use any combination of one or more readable media. The readable medium can be a readable signal medium or a readable storage medium. The readable storage medium can be, for example, but not limited to, an electrical, magnetic, optical, electromagnetic, infrared, or semiconductor system, device, or device, or any combination of the above. More specific examples of readable storage media (a non-exhaustive list) include: an electrical connection with one or more wires, a portable disk, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber, a portable compact disc read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the above.

[0239] A computer-readable storage medium may include a data signal transmitted in baseband or as part of a carrier wave, carrying readable program code. This transmitted data signal may take a variety of forms, including but not limited to electromagnetic signals, optical signals, or any suitable combination thereof. The readable storage medium may also be any readable medium other than a readable storage medium, which can transmit, transmit, or transfer a program for use by or in conjunction with an instruction execution system, apparatus, or device. The program code contained on the readable storage medium may be transmitted using any suitable medium, including but not limited to wireless, wired, optical cable, RF, etc., or any suitable combination thereof. The program code for performing the operations of the present invention may be written in any combination of one or more programming languages, including object-oriented programming languages ​​such as Java, C++, etc., as well as conventional procedural programming languages ​​such as "C" or similar programming languages. The program code may be executed entirely on the user computing device, partially on the user device, as a standalone software package, partially on the user computing device and partially on a remote computing device, or entirely on a remote computing device or server. Where a remote computing device is involved, the remote computing device may be connected to the user computing device through any type of network, including a local area network (LAN) or a wide area network (WAN), or may be connected to an external computing device (e.g., through the Internet using an Internet service provider).

[0240] The basic principles, main features, and advantages of the present invention are shown and described above. Those skilled in the art should understand that the present invention is not limited to the above embodiments. The above embodiments and descriptions are merely illustrative of the principles of the present invention. Various changes and modifications may be made to the present invention without departing from the spirit and scope of the present invention. Such changes and modifications are intended to fall within the scope of the present invention. The scope of protection claimed in the present invention is defined by the appended claims and their equivalents.

[0241] The present invention is explained from the perspectives of purpose of use, effectiveness, progress and novelty. The practical progress it has is in compliance with the functional enhancement and use requirements emphasized by the Patent Law. The above description and drawings of this application are only preferred embodiments of this application and are not intended to limit this application. Therefore, all structures, devices, features, etc. that are similar or identical to those of this application, that is, all equivalent replacements or modifications made in accordance with the scope of this patent application, should fall within the scope of protection of this patent application.

[0242] The specific implementation methods described above further illustrate the objectives, technical solutions and beneficial effects of the present invention in detail. It should be understood that the above description is only a specific implementation method of the present invention and is not intended to limit the scope of protection of the present invention. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principles of the present invention should be included in the scope of protection of the present invention.

Claims

1. A logistics vehicle carbon emission optimization method based on multi-objective collaboration, characterized by: The following steps are involved: S1. Obtain a multi-source data set of logistics vehicles, a set of evaluation indicators of logistics vehicles, and a data set of logistics vehicle tasks, and perform data correction on the multi-source data set of logistics vehicles through a Kalman filter model; S2. Based on the calibrated multi-source logistics vehicle dataset, an anomaly detection model is constructed using the kernel density estimation algorithm and the local outlier factor algorithm. Anomaly detection is performed on the calibrated multi-source logistics vehicle dataset to obtain logistics vehicle anomaly data. S3. Based on the abnormal data of logistics vehicles and the set of logistics vehicle evaluation indicators, a multi-dimensional carbon efficiency characteristic model is constructed to obtain a carbon efficiency characteristic matrix; S4. Based on the carbon efficiency characteristic matrix, a multi-objective carbon efficiency optimization model is constructed, and the multi-objective carbon efficiency optimization model is solved using the improved NSGA-II algorithm to obtain a Pareto frontier solution set; S5. Based on the logistics vehicle task dataset, a dynamic weight scoring model is constructed to evaluate the Pareto frontier solution set, obtain the optimal Pareto solution set, and implement it as a logistics vehicle carbon emission optimization strategy; The multi-source data set of logistics vehicles includes vehicle load data, vehicle access control system data, vehicle fuel consumption data, vehicle transportation mileage data and vehicle basic operation data; The logistics vehicle evaluation index set is of the same data type as the logistics vehicle multi-source dataset, and is used to perform data evaluation on the logistics vehicle multi-source dataset; The logistics vehicle task data set includes task allocation data and business constraint data; The basic vehicle operation data includes vehicle type, maximum vehicle load, vehicle fuel type and vehicle carbon emission coefficient; The business constraint data includes business priority, real-time road condition data and vehicle maintenance data.

2. The method for optimizing carbon emissions of logistics vehicles based on multi-objective collaboration according to claim 1 is characterized in that: The step S2 comprises the following steps: S201. Based on the corrected logistics vehicle multi-source data set, a logistics vehicle multi-source data matrix is ​​constructed, and a kernel density estimation algorithm is used to construct a joint probability density function; S202. Solve the joint probability density function by the local outlier factor algorithm to calculate the local outlier factor of each data point in the multi-source data matrix of logistics vehicles; S203. Based on the obtained local outlier factor, the data points are classified as abnormal using the K-means clustering algorithm to obtain abnormal data of logistics vehicles.

3. The method for optimizing carbon emissions of logistics vehicles based on multi-objective collaboration according to claim 2 is characterized in that: The local outlier factor algorithm uses Manhattan distance as a spatial distance metric.

4. The method for optimizing carbon emissions of logistics vehicles based on multi-objective collaboration according to claim 1 is characterized in that: The step S3 comprises the following steps: S301. Based on the abnormal data of logistics vehicles and the set of logistics vehicle evaluation indicators, a multi-dimensional carbon efficiency characteristic model is constructed; S302. Based on the multi-dimensional carbon efficiency feature model, perform feature association analysis using the DBSCAN clustering algorithm to obtain a carbon efficiency feature matrix.

5. A logistics vehicle carbon emission optimization system based on multi-objective collaboration, which is implemented based on a logistics vehicle carbon emission optimization method based on multi-objective collaboration as described in any one of claims 1 to 4, and is characterized in that: It includes a multi-source data acquisition module, anomaly detection module, multi-dimensional carbon efficiency feature extraction module, multi-objective dynamic optimization module and dynamic decision support module with sequential signal connection; The multi-source data acquisition module is used to obtain a multi-source data set of logistics vehicles, a logistics vehicle evaluation index set and a logistics vehicle task data set, and perform data correction on the multi-source data set of logistics vehicles through a Kalman filter model; The anomaly detection module is used to construct an anomaly detection model based on the corrected logistics vehicle multi-source data set through the kernel density estimation algorithm and the local interest group factor algorithm, perform anomaly detection on the corrected logistics vehicle multi-source data set, and obtain logistics vehicle anomaly data; The multi-dimensional carbon efficiency feature extraction module is used to construct a multi-dimensional carbon efficiency feature model based on the abnormal data of logistics vehicles and the logistics vehicle evaluation index set to obtain a structured carbon efficiency feature matrix; The multi-objective dynamic optimization module is used to construct a multi-objective carbon efficiency optimization model based on the structured carbon efficiency feature matrix, and solve the multi-objective carbon efficiency optimization model using the improved NSGA-II algorithm to obtain a Pareto frontier solution set; The dynamic decision support module is used to construct a dynamic weight scoring model based on the logistics vehicle task data set to evaluate the Pareto frontier solution set, obtain the optimal Pareto solution set, and implement it as a logistics vehicle carbon emission optimization strategy.

6. The multi-objective collaborative logistics vehicle carbon emission optimization system according to claim 5 is characterized in that: The system also includes a visualization module, which is connected to the dynamic decision support module signal for real-time display of logistics vehicle multi-source data sets, Logistics vehicle evaluation index set and logistics vehicle task data set, and real-time display of carbon emission detection results and optimization solutions.

7. A computer device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein: When the processor executes the computer program, it implements the logistics vehicle carbon emission optimization method based on multi-objective collaboration as described in any one of claims 1 to 4.

8. A storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, it implements the logistics vehicle carbon emission optimization method based on multi-objective collaboration as described in any one of claims 1 to 4.

Images