Intelligent distribution node carbon emission evaluation and optimization method

Abstract

The invention relates to the field of node carbon emission evaluation and optimization, in particular to an intelligent distribution node carbon emission evaluation and optimization method. The method comprises the following steps: firstly, preprocessing acquired operation data of a delivery node to obtain preprocessed operation data; on the basis of the preprocessed operation data, energy state vectors of the delivery nodes are obtained, an energy state flow configuration diagram is constructed, and energy state distortion is calculated; then, based on the energy state distortion amount, introducing a harmonic potential energy calculation formula to obtain harmonic potential energy; calculating a state-action cost based on the harmonic potential energy; calculating a total potential energy value based on the state-action cost; and finally, based on the total potential energy value, obtaining an optimal distribution path. The technical problems that the carbon emission level of the distribution node cannot be accurately evaluated, the carbon emission risk difference caused by energy consumption, load and traffic state between distribution paths cannot be quantified, and multi-dimensional harmonic optimization among carbon emission, distribution efficiency and distribution node load cannot be realized are solved.

Description

Technical Field

[0001] This invention relates to the field of node carbon emission assessment and optimization, and in particular to a method for assessing and optimizing carbon emissions from intelligent delivery nodes. Background Technology

[0002] With the continuous growth of urban logistics demand, intelligent delivery systems have become a key support for ensuring last-mile service quality and urban operational efficiency. However, existing delivery scheduling systems generally adopt scheduling strategies aimed at the shortest delivery route or optimal time, neglecting the carbon emission behavior during the delivery network operation. Especially in the context of widespread deployment of multi-energy hybrid vehicles and significant differences in carbon intensity among different types of delivery nodes, traditional scheduling mechanisms struggle to reflect the actual relationship between carbon emission levels and delivery route selection. Currently, existing delivery route optimization algorithms attempt to incorporate carbon emission factors into delivery route evaluation, but most use static carbon factor weighting, lacking the ability to dynamically model multi-source data such as the operational status of delivery nodes, load fluctuations, traffic congestion, and external disturbances. This results in significant deviations between carbon emission assessment results and actual values, making them unsuitable for real-time optimization of the scheduling process. Furthermore, existing delivery route optimization algorithms often employ single-objective optimization or linear multi-objective weighting methods, making it difficult to reasonably control carbon emission costs while ensuring delivery timeliness and operational stability. This is particularly true in urban environments with complex delivery node structures and dynamically changing delivery route networks, where there is a lack of comprehensive consideration of differences in delivery node states and risk propagation effects during delivery route jumps. Therefore, there is an urgent need to propose a method for assessing and optimizing carbon emissions from intelligent delivery nodes to address the aforementioned issues. Summary of the Invention

[0003] This invention provides a method for assessing and optimizing carbon emissions at intelligent delivery nodes, addressing the technical problems of being unable to accurately assess the carbon emission levels of delivery nodes, being unable to quantify the differences in carbon emission risks caused by energy consumption, load, and traffic conditions between delivery routes, and being unable to achieve multi-dimensional harmonization and optimization among carbon emissions, delivery efficiency, and delivery node load.

[0004] The present invention provides a method for assessing and optimizing carbon emissions from intelligent delivery nodes, comprising the following steps: S1. Based on the acquired operational data of the delivery nodes, preprocess the data to obtain preprocessed operational data; based on the preprocessed operational data, obtain the energy state vector of the delivery nodes; based on the energy state vector of the delivery nodes, construct an energy state flow configuration diagram and calculate the energy state distortion. S2. Based on the preprocessed operational data, the energy state vectors and energy state distortions of the delivery nodes, the harmonic potential energy calculation formula is introduced to obtain the harmonic potential energy; based on the harmonic potential energy, the state-action cost is calculated; based on the state-action cost, the total potential energy value is calculated; based on the total potential energy value, the optimal delivery path is obtained.

[0005] Preferably, S1 specifically includes: Based on the energy state vectors of delivery nodes, an energy state flow configuration graph is constructed, with delivery nodes serving as nodes in the energy state flow configuration graph and delivery paths between delivery nodes serving as edges in the energy state flow configuration graph.

[0006] Preferably, S1 specifically includes: Based on the energy state vector of the delivery node, the perturbation-normalized energy state difference is calculated; based on the perturbation-normalized energy state difference, the regularized envelope function term is calculated to obtain the energy state distortion variable of the delivery path.

[0007] Preferably, S2 specifically includes: Based on the preprocessed runtime data, the delivery route time is calculated, and combined with the obtained average speed of the delivery route, a time-speed coupling term is constructed.

[0008] Preferably, S2 specifically includes: The harmonic potential energy is calculated based on the time-velocity coupling term, the energy state distortion of the delivery path, and the load rate in the energy state vector of the delivery node.

[0009] Preferably, S2 specifically includes: The delivery path from the starting delivery node to the ending delivery node is combined into a complete delivery path chain; the external disturbance potential energy is calculated based on the energy state vector of the delivery node.

[0010] Preferably, S2 specifically includes: Based on harmonic potential energy and external disturbance potential energy, an operational deviation correction term is introduced to calculate the state-action cost.

[0011] Preferably, S2 specifically includes: The state-action costs are accumulated to obtain the total potential energy value of the complete delivery path chain; the complete delivery path chain corresponding to the minimum total potential energy value is taken as the optimal delivery path.

[0012] The beneficial effects of the technical solution of the present invention are: 1. By extracting the energy state vector from the preprocessed operational data and calculating the normalized carbon emission intensity in combination with the carbon emission conversion coefficient, not only is a quantitative assessment of the carbon emission level of each delivery node achieved, but also, through maximum and minimum value normalization processing, different types of delivery nodes, such as warehouses, transfer centers, and last-mile delivery points, are comparable under a unified scale. This effectively supports the perception and dynamic response to the carbon load status of delivery nodes, and improves the precision and dynamic adaptability of carbon emission assessment of delivery nodes.

[0013] 2. By constructing the delivery node state through a seven-dimensional energy state vector, introducing energy state distortion variables, and calculating the coupling effect between the energy state differences between delivery nodes and the disturbance normalized energy state difference degree and the path historical delay frequency, a nonlinear delivery path risk score is obtained. This enables delivery path scheduling to no longer make decisions based solely on time or distance, but to simultaneously consider the system instability caused by changes in carbon emissions, thereby avoiding sudden changes in carbon load caused by delivery path selection. This solves the problem that traditional delivery path scheduling models fail to consider the nonlinear coupling between carbon emission differences and delivery path risks.

[0014] 3. By integrating three key factors—carbon emission risk, time efficiency, and delivery node load pressure—into a single evaluation formula through the harmonic potential energy calculation formula, a consistent quantitative model of the comprehensive evaluation value of the delivery route is formed. This solves the problem that traditional scheduling systems have difficulty in uniformly modeling multi-dimensional factors and realizes a dynamic harmonization evaluation mechanism among carbon emission risk, time efficiency, and delivery node load pressure. Attached Figure Description

[0015] Figure 1 This is a flowchart of a method for assessing and optimizing carbon emissions from intelligent delivery nodes as described in this invention. Detailed Implementation

[0016] To further illustrate the technical means and effects adopted by the present invention to achieve its intended purpose, the technical solutions of the present invention will be clearly and completely described below with reference to the accompanying drawings of the embodiments. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0017] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention pertains.

[0018] The following description, in conjunction with the accompanying drawings, details a specific scheme for the intelligent delivery node carbon emission assessment and optimization method provided by the present invention.

[0019] See attached document Figure 1 The diagram illustrates a flowchart of a method for assessing and optimizing carbon emissions from a smart delivery node, provided by an embodiment of the present invention. The method includes the following steps: S1. Based on the acquired operational data of the delivery nodes, preprocess the data to obtain preprocessed operational data; based on the preprocessed operational data, obtain the energy state vector of the delivery nodes; based on the energy state vector of the delivery nodes, construct an energy state flow configuration diagram and calculate the energy state distortion.

[0020] The system acquires operational data from delivery nodes within the actual delivery platform, such as warehouses, transfer centers, and last-mile delivery points. This operational data includes the average number of deliveries completed per unit time (indicating node busyness), average vehicle load mass (reflecting node scheduling load), vehicle energy consumption per unit time (reflecting electricity or fuel consumption), carbon emission conversion factor (related to fuel type), real-time vehicle speed, delivery route congestion index, weather, actual delivery orders, and the number of vehicles available for dispatch during the same period. Data sources include databases from vehicle terminals, dispatch center systems, traffic data service platforms, and meteorological service systems. The acquired operational data undergoes preprocessing, including denoising (e.g., moving average), cleaning (e.g., filling), outlier handling (e.g., Z-score), time alignment (e.g., time zone conversion), standardization (e.g., Z-score standardization), and normalization (e.g., maximum and minimum values). The preprocessing processes employ techniques well-known to those skilled in the art and will not be elaborated upon here.

[0021] Based on the preprocessed operational data, for any delivery node The energy state vector of a delivery node is defined. This energy state vector is not simply concatenated, but formed through a mapping function with clearly defined physical meaning. Specifically, it is represented using a seven-dimensional structure. ,in, Indicates delivery node The energy state vector, It is the value in the first energy state dimension, representing the delivery node. The normalized carbon emission intensity is used to measure the delivery node. The carbon load level during the scheduling process is obtained from the delivery nodes through preprocessed operational data. The vehicle energy consumption per unit time is multiplied by the corresponding carbon emission conversion factor and then divided by the time window determined by the technicians. After normalization by the maximum and minimum values, the normalized carbon emission intensity is obtained, thus realizing the carbon emission assessment of the delivery node.

[0022] It is the value in the second energy state dimension, representing the delivery node. The normalized traffic impedance is used to reflect the delivery node. The average traffic difficulty of all delivery routes is calculated. The traffic impedance of any delivery route at each delivery node is obtained by dividing the length of the delivery route by the speed of the delivery route. The speed information on all delivery routes is obtained by back-calculating the vehicle GPS trajectory data in the existing database. The traffic impedance of all delivery routes at each delivery node is summed to obtain the traffic impedance of each delivery node. The traffic impedance of all delivery nodes is summed and then divided by the number of delivery nodes to obtain the average traffic impedance of all delivery nodes. Further, after calculating the average traffic impedance of all delivery routes at each delivery node, it is normalized by dividing it by the average traffic impedance of all delivery nodes to determine the normalized traffic impedance.

[0023] It is the value in the third energy state dimension, representing the delivery node. The current load rate is defined as the delivery node. Average vehicle load weight and delivery nodes The ratio of the maximum carrying capacity of the dispatched vehicles; the maximum carrying capacity of the dispatched vehicles is obtained from the existing vehicle registration information.

[0024] It is the value in the fourth energy state dimension, representing the delivery node. The magnitude of supply and demand fluctuations indicates the delivery node The intensity of fluctuations between order demand and vehicle supply is used to reflect the uncertainty of operational pressure at delivery nodes; by analyzing the fluctuations during the cycle... Internal delivery nodes The standard deviation of the sequence formed by the difference between the actual number of delivery orders and the sequence of available vehicles in the same period is calculated, and the obtained standard deviation is used as the supply and demand fluctuation range.

[0025] It is the value in the fifth energy state dimension, representing the delivery node. The eccentricity represents the delivery node. The spatial relative position of a delivery node in the global delivery path, that is, the delivery path from the starting delivery node to the ending delivery node. The relative distance ratio is calculated by measuring the distance from the starting delivery node to the delivery node. The ratio of the total distance of the delivery route to the total distance of the entire delivery route from the starting delivery node to the ending delivery node is used to obtain the total distance of the delivery route.

[0026] It is the value in the sixth energy state dimension, representing the delivery node. The vehicle energy consumption prediction error is obtained by normalizing the sum of squared residuals between historical energy consumption and predicted energy consumption using a method such as maximum-minimum normalization. Historical energy consumption is obtained from an existing database, and predicted energy consumption is obtained using existing machine learning modeling methods such as linear regression.

[0027] It is the value in the seventh energy state dimension, representing the delivery node. The intensity of external disturbances is determined by converting structured data such as weather levels, temporary traffic control, and road closure probabilities provided by meteorological and transportation platforms into disturbance scores through existing fuzzy logic reasoning systems.

[0028] After obtaining the energy state vectors of all delivery nodes, an energy state flow configuration graph is constructed, where delivery nodes are the nodes of the energy state flow configuration graph, and delivery paths between any delivery nodes are the edges of the energy state flow configuration graph. At this point, delivery paths between delivery nodes are no longer considered simple connections, but are given the physical meaning of energy state distortion propagation. Point to delivery node Calculate the energy state distortion variables of the delivery route. The formation process of energy state distortion values ​​depends entirely on the energy state vector, and the calculation formula is: in, Indicates delivery node With delivery nodes Energy state distortions between them, i.e., delivery paths Energy state distortion; It is an index of the energy state dimension; and These represent delivery nodes. With delivery nodes The Values ​​in each energy state dimension; It is a one-dimensional energy state difference squared term used to describe delivery nodes. With delivery nodes In the The intensity of the difference in each energy state dimension; It is the first The weighting coefficients for each energy state dimension are determined using the existing entropy weighting method, with a reference range of values. And the sum is 1; It is the path disturbance factor, representing the delivery node. To delivery node The intensity of timeliness uncertainty along the delivery route is determined using existing real-time traffic condition assessment models, with a reference value range of [value range missing]. ; It is the path history delay frequency, used to describe the delivery node. To delivery node The probability level of a delivery route being delayed in its historical operation is determined by the ratio of the number of delays to the total number of passages for that delivery route segment retrieved from the existing database. It is a historical risk modulation term, used to smoothly introduce historical delayed risks in a logarithmic manner to avoid numerical explosion caused by high-frequency anomalies, while retaining the risk amplification effect. It is a weighted sum of squared differences in Euclidean energy states, used to describe delivery nodes. To delivery node The weighted squared Euclidean distance of the energy state vector differences; It is the perturbation normalized energy state difference degree, which reflects the tendency to reduce the sensitivity of energy state differences when the delivery route is occupied by instability; The differences in perturbation energy states after delay risk modulation are described; It is a regularized envelope function term used to enable the energy state distorted variable to become a delivery route risk score with a regular lower bound and a nonlinear upper bound, which can be directly used as a measure of delivery route energy consumption risk.

[0029] S2. Based on the preprocessed operational data, the energy state vectors and energy state distortions of the delivery nodes, the harmonic potential energy calculation formula is introduced to obtain the harmonic potential energy; based on the harmonic potential energy, the state-action cost is calculated; based on the state-action cost, the total potential energy value is calculated; based on the total potential energy value, the optimal delivery path is obtained.

[0030] After calculating the energy state distortions for all delivery routes, the selection of delivery routes is not performed immediately. Instead, the energy state distortions are further incorporated into the harmonic potential energy calculation formula. The energy state distortions, time consumption, and load rates of delivery nodes for each delivery route are uniformly included in the harmonic potential energy calculation formula. Specifically, firstly, based on real-time traffic data from the preprocessed operational data, such as real-time vehicle speed and traffic flow, the existing traffic prediction system is used to calculate the delivery routes. time And combined with delivery routes obtained through vehicle GPS trajectory data average speed Construct time-velocity coupling terms; simultaneously, through the delivery nodes in the energy state vector. Current load rate A nonlinear amplification term is introduced to characterize the nonlinear growth of carbon emission risk under high load conditions. The final formula for calculating the harmonic potential energy is: in, Delivery route The harmonic potential energy is used to represent the comprehensive evaluation value of the combined effects of three factors, namely, "carbon emission risk, time efficiency, and delivery node load pressure," under the current delivery route segment. This is the energy state distortion weighting factor, used to represent the level of concern regarding the risk of energy state disturbances along the delivery route. It is determined using existing regression analysis methods based on historical route energy state distortion values ​​and historical carbon emission increments obtained from existing databases. The reference value range is [insert range here]. ; This is a time-velocity harmonic weighting factor used to reflect the sensitivity to the hidden carbon risks caused by efficiency degradation. It is determined through regression analysis based on historical route data and vehicle carbon emission records from an existing database, with a reference value range of [value missing]. ; It is the maximum reference travel time, that is, the maximum travel time among all delivery routes; It is the average speed of the maximum path, that is, the average speed of the delivery path corresponding to the maximum travel time; This is a load harmonization weighting factor, used to represent the control intensity of load pressure on delivery nodes. It is determined using polynomial fitting or spline regression based on historical delivery node load rates and carbon emission conversion factors obtained from existing databases. The reference value range is... ; Indicates delivery node The current load rate; Describes the process from the delivery node to delivery node Corresponding delivery route The degree of energy state distortion; It is a time-velocity coupling term used to express the delivery route. The coupling effect between relative time consumption and relative traffic efficiency; It is a nonlinear amplification term used to characterize the impact of the load saturation level of delivery nodes on carbon emissions and transportation capacity safety.

[0031] After obtaining the harmonic potential energy, to achieve dynamic scheduling optimization across the entire delivery route, delivery nodes are treated as nodes in a recursive scheduling decision tree. This tree is constructed by treating each possible combination of delivery routes as a state evolution chain, i.e., a complete delivery route chain, and the local state jump cost of this chain is calculated based on the harmonic potential energy. This process is repeated at each step. Calculate once per cycle, outputting the harmonic potential energy. This serves as the energy state cost for the current action, and the state-action cost is evaluated based on the energy state cost. This is used for delivery path search and pruning decisions in recursive scheduling decision trees. The formula for calculating state-action cost is: in, It is in delivery status. Next action The state-action cost, i.e. the local state jump cost of the state evolution chain, is used to represent the immediate scheduling cost of selecting a certain delivery path in the current delivery state. The smaller the value, the better. Indicates at time step The delivery status includes the current delivery node and historical information about the delivery route, such as the time step. Delivery status is At this time, it indicates that the vehicle is currently located at delivery node 3, and has visited 2 delivery route segments, meaning it has experienced 2 delivery status transitions. In time step Action variables, Describes the time step From delivery nodes Heading to delivery node The action of selecting the delivery route; This is the scheduling value scaling factor, used to adjust the rating range for different delivery routes to prevent distorted scheduling due to excessively large or small values. The standard deviation of historical ratings obtained from the existing database is used as the scheduling value scaling factor, with a reference value range of [missing value]. ; This is the perturbation weight coefficient, used to control the proportion of external environmental factors in the scheduling score. It is determined using the existing Bayesian optimization method, and the reference value range is [value range missing]. ; It is the external disturbance potential energy, used to represent the potential energy from the delivery node. to delivery node The additional potential energy caused by external environmental disturbances , It is a delivery node The eccentricity, For delivery nodes The intensity of external disturbances, It is the path history delay frequency; It is a running deviation correction term used to represent the delivery route. Error accumulation from historical executions is used to penalize unstable delivery routes, based on delivery routes obtained from the existing database. The corresponding historical deviation data between the plan and the actual data were obtained by fitting the data using the least squares method. The deviation data between the plan and the actual data included time overruns, abnormal carbon emissions, and detour frequency. It is a harmonic term used to represent the balance loss assessment between the theoretical optimum and environmental uncertainty of the delivery route; It is a disturbance risk weighting factor used to reflect the environmental awareness capability in delivery route selection; It is the combined scheduling risk value, used to reflect the comprehensive scheduling cost of delivery routes.

[0032] Furthermore, in the recursive scheduling decision tree, starting from the root node (i.e., the initial delivery node), the actions of each delivery path are evaluated sequentially. The state-action cost of all delivery path actions is accumulated to form a complete delivery path chain. Total potential energy , This represents a complete delivery route chain, such as 0. .

[0033] Finally, based on all feasible complete delivery route chains, the optimal delivery route is selected. This enables the optimization of intelligent delivery nodes.

[0034] In summary, a method for assessing and optimizing carbon emissions from intelligent delivery nodes has been developed.

[0035] The order of the embodiments is for illustrative purposes only and does not represent the superiority or inferiority of the embodiments. The processes depicted in the drawings do not necessarily require a specific or sequential order to achieve the desired result. In some embodiments, multitasking and parallel processing are possible or may be advantageous.

[0036] The various embodiments in this specification are described in a progressive manner. The same or similar parts between the various embodiments can be referred to each other. Each embodiment focuses on describing the differences from other embodiments.

[0037] The above embodiments are only used to illustrate the technical solutions of the present invention, and are not intended to limit it. Although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features. Such modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of the present invention, and should all be included within the protection scope of the present invention.

Claims

1. A method for assessing and optimizing carbon emissions from intelligent delivery nodes, characterized in that, Includes the following steps: S1. Based on the acquired operational data of the delivery nodes, preprocess the data to obtain preprocessed operational data; Based on the preprocessed operational data, the energy state vector of the delivery node is obtained; Based on the energy state vectors of delivery nodes, an energy state flow configuration diagram is constructed, and energy state distortion variables are calculated. S2. Based on the preprocessed operational data, the energy state vectors and energy state distortions of the delivery nodes, a harmonic potential energy calculation formula is introduced to obtain the harmonic potential energy; Calculate the state-action cost based on harmonic potential energy; Calculate the total potential energy value based on the state-action cost; The optimal delivery route is obtained based on the total potential energy value.

2. The method for assessing and optimizing carbon emissions from intelligent delivery nodes according to claim 1, characterized in that, S1 specifically includes: Based on the energy state vectors of delivery nodes, an energy state flow configuration graph is constructed, with delivery nodes serving as nodes in the energy state flow configuration graph and delivery paths between delivery nodes serving as edges in the energy state flow configuration graph.

3. The method for assessing and optimizing carbon emissions from intelligent delivery nodes according to claim 2, characterized in that, S1 specifically includes: Based on the energy state vector of the delivery node, the perturbation-normalized energy state difference is calculated; based on the perturbation-normalized energy state difference, the regularized envelope function term is calculated to obtain the energy state distortion variable of the delivery path.

4. The method for assessing and optimizing carbon emissions from intelligent delivery nodes according to claim 1, characterized in that, S2 specifically includes: Based on the preprocessed runtime data, the delivery route time is calculated, and combined with the obtained average speed of the delivery route, a time-speed coupling term is constructed.

5. The method for assessing and optimizing carbon emissions from intelligent delivery nodes according to claim 4, characterized in that, S2 specifically includes: The harmonic potential energy is calculated based on the time-velocity coupling term, the energy state distortion of the delivery path, and the load rate in the energy state vector of the delivery node.

6. The intelligent distribution node carbon emission evaluation and optimization method according to claim 5, characterized in that, S2 specifically includes: The delivery path from the starting delivery node to the ending delivery node is combined into a complete delivery path chain; the external disturbance potential energy is calculated based on the energy state vector of the delivery node.

7. The intelligent distribution node carbon emission evaluation and optimization method of claim 6, wherein, S2 specifically includes: Based on harmonic potential energy and external disturbance potential energy, an operational deviation correction term is introduced to calculate the state-action cost. 8.The intelligent distribution node carbon emission evaluation and optimization method of claim 7, wherein, S2 specifically includes: The state-action costs are accumulated to obtain the total potential energy value of the complete delivery path chain; the complete delivery path chain corresponding to the minimum total potential energy value is taken as the optimal delivery path.

Images