Terminal delivery planning method and system

Abstract

The present invention relates to the field of path planning technology, and specifically to a terminal delivery planning method and system, comprising the following steps: S1, real-time acquisition of basic traffic data, traffic control data, and customer timeliness demand data for each road section in the target area, generating a dynamic road network state matrix; S2, calculating the spatiotemporal integrated priority weight value of each delivery node based on the dynamic road network state matrix, generating a priority delivery scheduling sequence; S3, based on the priority delivery scheduling sequence, using a dynamic window algorithm to generate an optimal delivery path. The present invention effectively alleviates distribution resource conflicts in high-density delivery scenarios, improves overall service stability and scheduling rationality, achieves rapid path convergence and scheduling adaptation in a real-time environment, and significantly improves overall performance in terms of energy consumption control and delivery success rate compared to traditional fixed strategies.

Description

Technical Field

[0001] The present invention relates to the field of route planning technology, and in particular to a terminal delivery planning method and system. Background Art

[0002] With the rapid growth of urban logistics demand, terminal delivery tasks are facing increasingly severe challenges in timeliness, energy consumption control and route intelligence. Existing delivery route planning systems are mostly based on static map information and a single weight indicator (such as shortest distance or optimal time) for route selection. They lack the ability to jointly model the dynamic state of complex urban traffic, customers' diverse time requirements and multi-objective trade-off strategies. As a result, delivery systems are prone to route conflicts, energy waste and time limit violations in high-density and high-frequency scenarios.

[0003] Some existing methods incorporate time windows or traffic flow parameters as optimization references, but these methods often overlook the real-time fluctuations in path capacity, failing to construct a complete dynamic road network state matrix. This makes it difficult to accurately predict the impact of disruptions or congestion on subsequent deliveries. Furthermore, in the process of sorting delivery nodes, static priority rules or simple time urgency judgments are often used, lacking comprehensive consideration of spatial distribution, concentration, and accessibility, making it impossible to form an effective hierarchical delivery management mechanism.

[0004] In addition, most existing path optimization algorithms use fixed windows, single-objective optimization or heuristic greedy strategies, which fail to respond promptly when faced with sudden situations such as dynamic order insertion, power constraints or local high-density delivery. This results in high rigidity and poor flexibility in path planning, and a lack of a multi-objective weight mechanism based on adaptive adjustment of operating status. Summary of the Invention

[0005] The present invention provides a terminal delivery planning method, which improves the timeliness reliability, energy economy and intelligent scheduling level of the terminal delivery system through traffic dynamic modeling, spatiotemporal joint sorting and rolling adjustable multi-objective path optimization method.

[0006] A terminal delivery planning method includes the following steps:

[0007] S1. Obtaining in real time basic traffic data, traffic control data, and customer time efficiency demand data for each road section in the target area, and generating a dynamic road network status matrix. The dynamic road network status matrix includes basic travel time indexed by road section number, real-time traffic fluctuation coefficient, and road section connection status identifier;

[0008] S2. Calculate the spatiotemporal integrated priority weight value of each dispatch node based on the dynamic road network state matrix to generate a priority dispatch scheduling sequence, wherein the spatiotemporal integrated priority weight value is calculated by coupling the timeliness sensitivity and spatial proximity index of the dispatch node through a calculation function;

[0009] S3. Based on the priority delivery scheduling sequence, a dynamic window algorithm is used to generate an optimal delivery path. The dynamic window algorithm uses a dual constraint mechanism based on a service time window and an energy consumption window to optimize the total energy consumption and response time of the path.

[0010] Optionally, in S1:

[0011] The basic traffic data is periodically acquired through the traffic information platform interface, including the design speed of the road section, the historical average travel time, the number of lanes and the phase data of the signal light, and is stored in the road network database according to the road section number;

[0012] The traffic control data is accessed through the real-time data stream of the traffic control system, and the controlled road section number, control type and duration are analyzed to mark the temporarily controlled road section as abnormally connected, and the traffic capacity attenuation coefficient of the controlled period is generated;

[0013] The customer timeliness requirement data is extracted from the order management system, including the customer's geographic location coordinates, promised delivery time window and expedited identifier, and a demand feature vector is established according to the delivery node number.

[0014] Optionally, the capacity attenuation coefficient during the control period is expressed as: ,in, Indicates time The capacity attenuation coefficient is is the road section carrying capacity factor, It is the control level, and the level value is obtained by mapping the control type.

[0015] Optionally, the dynamic road network state matrix is ​​constructed using the road segment number as the index, and the tuple corresponding to each road segment is represented as:

[0016] ,in:

[0017] Indicates basic travel time: ,in, For road sections length, For road sections The historical average vehicle speed during the current period, is the weather correction factor;

[0018] Indicates the real-time traffic fluctuation coefficient: ,in, For road sections The design speed, For road sections At the moment The actual vehicle speed calculated by the floating vehicle;

[0019] Indicates the link connection status: ,in, Indicates that the road is clear. Indicates slow traffic on the road. Indicates that the road section is interrupted;

[0020] Indicates road section At the moment The degree of capacity loss due to traffic control, with a value range of 0 to 1. A larger value indicates a more serious decline in capacity.

[0021] Optionally, the S1 also includes a linkage update mechanism, which updates the dynamic road network state matrix once at a preset time interval. , that is, when it is interrupted, the fluctuation coefficients of adjacent sections are triggered to be updated in a linked manner.

[0022] Optionally, the S2 specifically includes:

[0023] S21: For each delivery node, a time-sensitive coefficient with a sudden change characteristic is constructed based on the time difference between the latest delivery time and the current system time, combined with a preset emergency response threshold. This time-sensitive coefficient is used to characterize the time urgency of the order corresponding to the delivery node, and the time-sensitive adjustment factor is used to adaptively reflect the total pressure of undelivered tasks in the current system.

[0024] S22: Extract the shortest accessible path from the current vehicle location to each delivery node from the dynamic road network state matrix and calculate the actual travel time for the corresponding path. Simultaneously, a fixed radius is set around each delivery node. The number of undelivered nodes within the radius is counted and combined with their spatial distribution variance to form a spatial cluster density. Potential high-density delivery areas are identified. A weighted synthesis of path accessibility and spatial cluster density is then performed to construct a complete spatial proximity index.

[0025] S23, by integrating the time sensitivity coefficient and the spatial proximity index, constructs the spatiotemporal comprehensive priority weight value of the delivery node. After normalization, it is divided into levels according to the dual threshold strategy to form high priority, medium priority and delayable delivery subsequences respectively, and sets sorting rules according to the spatiotemporal comprehensive priority weight value or spatial aggregation to construct a global priority delivery scheduling sequence.

[0026] Optionally, the S2 further includes a dynamic adjustment mechanism, which includes: if a new delivery task is detected or the road network status changes, recalculating the indicators of all nodes that have not completed the delivery.

[0027] Optionally, the constraints of the service time window include: calculating the latest service time based on the promised delivery time of the delivery node and the current system time, and setting an adaptive buffer as the lower limit of the service time;

[0028] The constraints of the energy consumption window include: constructing a unit distance energy consumption prediction model based on the current vehicle load and speed, and estimating the available energy consumption budget in combination with the remaining power to ensure that the generated path meets the dual conditions of timeliness and endurance. The unit distance energy consumption prediction model is a quadratic function of load and speed, including vehicle power coefficient, linear resistance coefficient and static energy consumption terms; the available energy consumption budget is the available energy consumption after deducting the consumed amount from the total energy and retaining the safety threshold.

[0029] Optionally, the S3 specifically includes:

[0030] S31, Window Initialization: A fixed number of dispatch nodes are intercepted from the priority dispatch scheduling sequence to form a sliding window, and all arranged paths are filtered for feasibility. If a path includes a section with a disconnected connection status or does not meet the service time window and energy consumption window constraints, it is eliminated, and only the paths that meet the constraints are retained and enter the optimization process;

[0031] S32, Path Optimization: A multi-objective evaluation is used to score feasible paths. The objective function simultaneously considers the total path time, cumulative energy consumption, and node spatial clustering density, and performs weighted fusion through adjustable weight coefficients. Among them, the time constraint is a rigid constraint, and the energy consumption allows a relaxation ratio to be set within the budget. The ε-constraint method is used to search for non-dominated solution sets, and the comprehensive optimal path is selected based on the objective function value.

[0032] A terminal delivery planning system, used to implement the above-mentioned terminal delivery planning method, includes the following modules:

[0033] Dynamic road network data acquisition module, used to obtain basic traffic data, traffic control data and customer time efficiency demand data of each road section in the target area, and build a dynamic road network status matrix;

[0034] A dispatch node weight calculation module, configured to calculate the spatiotemporal integrated priority weight value of each dispatch node based on the dynamic road network state matrix, and generate a priority dispatch scheduling sequence;

[0035] The path optimization module is used to generate an optimal delivery path based on the priority delivery scheduling sequence by adopting a dynamic window algorithm based on the dual constraints of service time window and energy consumption window.

[0036] Beneficial effects of the present invention:

[0037] The present invention improves the sensitivity of distinguishing orders with critical time limits by constructing a spatiotemporal weight model that integrates "time sensitivity" and "spatial proximity". At the same time, it dynamically evaluates the value of nodes by combining path accessibility and delivery point concentration. The multi-level priority queue division mechanism significantly reduces the risk of time default caused by misjudgment, and effectively alleviates distribution resource conflicts in high-density delivery scenarios, thereby improving overall service stability and scheduling rationality.

[0038] The present invention introduces the dual constraints of time window and energy consumption budget, generates a feasible path set based on the full arrangement of nodes in the sliding window, and performs global scoring by fusing path time, unit energy consumption and node cluster density through a multi-objective function; on this basis, the ε-constraint method is used to construct a non-dominated solution set, achieving rapid path convergence and scheduling adaptation in a real-time environment, and significantly improving the overall performance in terms of energy consumption control and delivery success rate compared with traditional fixed strategies.

[0039] The present invention adjusts the three types of weight ratios in the path optimization objective function in real time according to the delivery task status, power level and local node density, so that the system can automatically switch strategy priorities in typical scenarios such as emergency order insertion, endurance warning and high-density delivery, and maintain collaborative optimization among multiple objectives. BRIEF DESCRIPTION OF THE DRAWINGS

[0040] In order to more clearly illustrate the technical solutions in the present invention or the prior art, the following briefly introduces the drawings required for use in the embodiments or the description of the prior art. Obviously, the drawings described below are only for the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative work.

[0041] Figure 1 A schematic diagram of a planning method flow chart according to an embodiment of the present invention;

[0042] Figure 2 Schematic diagram of the functional modules of the planning system according to an embodiment of the present invention. DETAILED DESCRIPTION

[0043] The present invention is described in detail below with reference to the accompanying drawings and specific embodiments. It is also noted that, to provide a more detailed description, the following embodiments are best and preferred embodiments, and those skilled in the art may employ alternative methods for implementing certain known technologies. Furthermore, the accompanying drawings are intended only to provide a more detailed description of the embodiments and are not intended to limit the present invention.

[0044] It should be noted that references in the specification to "one embodiment," "an embodiment," "exemplary embodiments," "some embodiments," etc. indicate that the described embodiments may include specific features, structures, or characteristics, but not necessarily every embodiment will include such specific features, structures, or characteristics. Furthermore, when specific features, structures, or characteristics are described in conjunction with an embodiment, it is within the knowledge of persons skilled in the relevant art to implement such features, structures, or characteristics in conjunction with other embodiments (whether or not explicitly described).

[0045] In general, terms can be understood, at least in part, from their use in context. For example, depending at least in part on the context, the term "one or more" as used herein can be used to describe any feature, structure, or characteristic in the singular sense, or can be used to describe a combination of features, structures, or characteristics in the plural sense. Additionally, the term "based on" can be understood as not necessarily intended to convey an exclusive set of factors, but can instead, depending at least in part on the context, allow for the presence of other factors that are not necessarily explicitly described.

[0046] like Figure 1 As shown, a terminal delivery planning method includes the following steps:

[0047] S1. Obtain basic traffic data, traffic control data, and customer time efficiency demand data for each road section in the target area in real time, and generate a dynamic road network status matrix. The dynamic road network status matrix includes basic travel time indexed by road section number, real-time traffic fluctuation coefficient, and road section connection status identification;

[0048] S2. Calculate the spatiotemporal integrated priority weights of each dispatch node based on the dynamic road network state matrix to generate a priority dispatch scheduling sequence. The spatiotemporal integrated priority weights are calculated using a coupling function based on the timeliness sensitivity and spatial proximity index of the dispatch node.

[0049] S3. Based on the priority dispatch scheduling sequence, a dynamic window algorithm is used to generate the optimal dispatch path. The dynamic window algorithm adopts a dual constraint mechanism based on the service time window and the energy consumption window to optimize the total energy consumption and response time of the path.

[0050] S1 specifically includes:

[0051] S11, periodically obtain basic traffic data of the road section through the traffic information platform interface, including the design speed of the road section, historical average travel time, number of lanes and signal light phase data, and store it in the road network database according to the road section number;

[0052] S12: Access the real-time data stream of the traffic control system, analyze the controlled section number, control type, and duration, mark the temporarily controlled section as abnormally connected, and generate the capacity attenuation coefficient for the controlled period: ,in, Indicates time The capacity attenuation coefficient is is the road section carrying capacity factor, The control level (the level value obtained by mapping the control type) is converted into an integer level value through the preset mapping rules, as follows:

[0053] Level 1 (G=1): Slight restrictions, such as temporary construction bypass, one-side closure but passable;

[0054] Level 2 (G=2): Moderate restrictions, such as half-width closure and time-limited one-way traffic;

[0055] Level 3 (G=3): Severe restrictions, such as total closure, accident blockade, and traffic control;

[0056] The larger the level value, the greater the impact on traffic capacity and the higher the corresponding attenuation coefficient.

[0057] S13, extract customer timeliness requirement data from the order management system, including customer geographic location coordinates and promised delivery time window and expedited identifier, and establish a demand feature vector according to the dispatch node number;

[0058] S14, construct a dynamic road network state matrix, with the road section number is the index, and the tuple corresponding to each road segment is represented as:

[0059] ;

[0060] The components are defined as follows:

[0061] Basic travel time: ,in, For road sections length, For road sections The historical average vehicle speed during the current period, is the weather correction factor (set according to meteorological data, the range is ;

[0062] Real-time traffic fluctuation coefficient: ,in, For road sections The design speed, For road sections At the moment The actual vehicle speed calculated by the floating vehicle;

[0063] Link connection status indicator: ,in, Indicates that the road is clear. Indicates slow traffic on the road. Indicates road interruption (closure or severe congestion);

[0064] S15, update the matrix data every 5 minutes. When the link is interrupted, the fluctuation coefficient linkage update mechanism of the adjacent sections is triggered:

[0065] ,in, Represents and interrupts the road segment Adjacent road segment numbers, Indicates a new entry point due to diversion The number of vehicles entering the vehicle (which can be estimated through entrance counts or predictive models).

[0066] S2 specifically includes:

[0067] S21, Time Sensitivity Calculation: For Delivery Nodes , calculate its time sensitivity coefficient:

[0068] ;

[0069] in, Representation node The latest promised delivery time, Indicates the current system time. The emergency response threshold is 30 minutes by default. is the time sensitivity adjustment factor, which is dynamically adjusted according to the total number of undelivered items ( , Indicates the total number of orders that have not been delivered at the current moment.

[0070] S22, spatial proximity index calculation:

[0071] S221, real-time travel time estimation:

[0072] ;in, Indicates the current vehicle position To the delivery node The shortest path contains a set of road segments, Indicates road section The basic travel time, Indicates road section At the moment Real-time traffic fluctuation coefficient;

[0073] S222, spatial aggregation density calculation:

[0074] ;in, Represents a node is the center of the circle, and the radius The number of undelivered nodes within the range, These nodes to The Euclidean distance variance reflects the degree of spatial concentration;

[0075] S223, spatial proximity index:

[0076] ;in, , is the spatial accessibility weight, which decreases linearly from 0.7 to 0.3 as the delivery progresses. It is the aggregation density weight, which supplements the spatial distribution impact.

[0077] ;in, , is the spatial accessibility weight, which decreases linearly from 0.7 to 0.3 as the delivery progresses. It is the aggregation density weight, which supplements the spatial distribution impact.

[0078] S23, coupled calculation and sequence generation:

[0079] S231, constructing the space-time coupling integrated priority weight value:

[0080] ;in, is the fusion weight factor, with an initial value of 0.6, and increases by 0.1 after completing 20% ​​of the delivery task, with a maximum value of 0.9. They are the minimum and maximum values ​​of the spatial index of all nodes to be dispatched at the current moment.

[0081] S232, threshold classification and queue division, according to Value-partitioned priority queue:

[0082] like ( ), it is marked as a red emergency node and added to the subsequence ;

[0083] like ( , then it is a yellow regular node, add subsequence ;

[0084] Otherwise, it is marked as a green deferrable node and added to the queue. ;

[0085] S233, global priority sequence generation rules: ;

[0086] subsequence Internal Press Sort in descending order;

[0087] subsequence 、 Internal Press Sort in ascending order;

[0088] S24, Dynamic Adjustment Mechanism: Each time a new delivery request is received or a network status update is detected, if the combined priority weight of the newly added node is greater than the lowest value of the current emergency subsequence, it will be inserted at the front of the sequence for priority processing. At the same time, for nodes approaching timeout, the system will forcibly increase the proportion of the timeliness component of their fusion weight, ensuring that the delivery strategy responds promptly to timeliness risks and implementing a dynamically adjustable delivery priority control mechanism. Specifically, the following actions are performed:

[0089] S241. Recalculate only the nodes that have not been dispatched , ;

[0090] S242. If the new node , then insert Team leader;

[0091] S243. If the remaining time of a node , then it is mandatory to set .

[0092] S3 specifically includes:

[0093] S31, Constraint Modeling:

[0094] S311, Time Window Constraint: Calculate the Delivery Node Latest service time:

[0095] ;

[0096] The buffer time is defined as: ;

[0097] Path feasibility conditions: ;

[0098] in, For nodes The promised delivery time, For nodes The current remaining available time, The first Estimated travel time for each road section, The current system time.

[0099] S312, energy consumption window constraint:

[0100] Energy consumption per unit distance prediction model: ;

[0101] Remaining energy budget: ;

[0102] in, is the current load, is the current average driving speed, is the empirical coefficient of the vehicle power system, is the total power consumption budget, is the energy consumed; To ensure the energy threshold is retained;

[0103] S32, dynamic window initialization:

[0104] S321, intercept the first one from the priority delivery sequence nodes constitute a sliding window ( ;

[0105] S322, perform full permutation and combination of the nodes in the window to construct a candidate path set.

[0106] S323, filter the path to exclude the path containing the connection status Impassable road sections;

[0107] S324: Keep the paths that meet the following conditions:

[0108] The total path time does not exceed the node's latest service time;

[0109] The cumulative unit energy consumption does not exceed the remaining budget.

[0110] S33, Multi-Objective Optimization:

[0111] S331, construct the objective function: ;in, It represents the value of the comprehensive path evaluation objective function. The smaller it is, the better the path is. , the initial value is 、 、 , is the time cost weight, which indicates the importance attached to the optimization of the total travel time. is the energy consumption cost weight, which indicates the importance attached to energy consumption control. is the weight for optimizing spatial aggregation, indicating the importance attached to intensive task delivery. For the dispatch node The spatial aggregation density index, represents the sum of the travel time of all road segments in the selected path, represents the total energy consumption per unit distance of each section in the selected path (multiplied by the distance), Represents the sum of the spatial clustering densities of each dispatch node in the selected path;

[0112] S332, optimization method adopts -Bound Law:

[0113] The time window is a rigid constraint;

[0114] The energy consumption budget is allowed to fluctuate within a range of 10%;

[0115] Search for non-dominated solutions within this range to form the Pareto front solution set.

[0116] S34, Path Decision and Rolling Optimization:

[0117] S341, select the comprehensive optimal path:

[0118] If there exists a minimum value that satisfies the objective function And the maximum aggregation density The path is selected first;

[0119] If there is no path that meets the above conditions, the lexicographic order decision rule is adopted, giving priority to meeting the time requirement and secondly optimizing energy consumption;

[0120] S342, perform rolling window update:

[0121] After each delivery of two nodes, the window automatically slides; three new undelivered nodes are loaded; the completed path segments are retained, and only the subsequent segments are replanned.

[0122] S35, dynamically adjust the weight ratio in the objective function according to the operating status:

[0123] If the system inserts an urgent order, the adjustment is: ;

[0124] If the remaining power alarm is detected (i.e. ), then adjust to: , appropriately reduce and ;

[0125] If the current path aggregation density is greater than the set threshold ( ), then the improvement Increase by 0.15.

[0126] like Figure 2 As shown, a terminal delivery planning system is used to implement the above planning method, including the following modules:

[0127] Dynamic road network data acquisition module, used to obtain basic traffic data, traffic control data and customer time efficiency demand data of each road section in the target area, and build a dynamic road network status matrix;

[0128] A dispatch node weight calculation module, configured to calculate the spatiotemporal integrated priority weight value of each dispatch node based on the dynamic road network state matrix, and generate a priority dispatch scheduling sequence;

[0129] The path optimization module is used to generate an optimal delivery path based on the priority delivery scheduling sequence by adopting a dynamic window algorithm based on the dual constraints of service time window and energy consumption window.

[0130] The present invention encompasses any alternatives, modifications, equivalents, and solutions that fall within the spirit and scope of the present invention. To provide a thorough understanding of the present invention, specific details are described in detail below in connection with the preferred embodiments of the present invention, but those skilled in the art will be able to fully understand the present invention without these detailed descriptions. Furthermore, to avoid unnecessary confusion regarding the essence of the present invention, well-known methods, processes, procedures, components, and circuits have not been described in detail.

[0131] The above is only a preferred embodiment of the present invention. It should be pointed out that for ordinary technicians in this technical field, several improvements and modifications can be made without departing from the principles of the present invention. These improvements and modifications should also be regarded as within the scope of protection of the present invention.

Claims

1. A terminal delivery planning method, characterized in that: The following steps are involved: S1. Obtaining in real time basic traffic data, traffic control data, and customer time efficiency demand data for each road section in the target area, and generating a dynamic road network status matrix. The dynamic road network status matrix includes basic travel time indexed by road section number, real-time traffic fluctuation coefficient, and road section connection status identifier; S2. Calculate the spatiotemporal integrated priority weight value of each dispatch node based on the dynamic road network state matrix to generate a priority dispatch scheduling sequence, wherein the spatiotemporal integrated priority weight value is calculated by coupling the timeliness sensitivity and spatial proximity index of the dispatch node through a calculation function; S3. Based on the priority delivery scheduling sequence, a dynamic window algorithm is used to generate an optimal delivery path. The dynamic window algorithm uses a dual constraint mechanism based on a service time window and an energy consumption window to optimize the total energy consumption and response time of the path.

2. A terminal delivery planning method according to claim 1, characterized in that: In S1: The basic traffic data is periodically acquired through the traffic information platform interface, including the design speed of the road section, the historical average travel time, the number of lanes and the phase data of the signal light, and is stored in the road network database according to the road section number; The traffic control data is accessed through the real-time data stream of the traffic control system, and the controlled road section number, control type and duration are analyzed to mark the temporarily controlled road section as abnormally connected, and the traffic capacity attenuation coefficient of the controlled period is generated; The customer timeliness requirement data is extracted from the order management system, including the customer's geographic location coordinates, promised delivery time window and expedited identifier, and a demand feature vector is established according to the delivery node number.

3. A terminal delivery planning method according to claim 2, characterized in that: The capacity attenuation coefficient during the control period is expressed as: ,in, Indicates time The capacity attenuation coefficient is is the road section carrying capacity factor, It is the control level, and the level value is obtained by mapping the control type.

4. A terminal delivery planning method according to claim 3, characterized in that: The construction of the dynamic road network state matrix is ​​indexed by the road section number, and the tuple corresponding to each road section is represented as: ,in: Indicates basic travel time: ,in, For road sections length, For road sections The historical average vehicle speed during the current period, is the weather correction factor; Indicates the real-time traffic fluctuation coefficient: ,in, For road sections The design speed, For road sections At the moment The actual vehicle speed calculated by the floating vehicle; Indicates the link connection status: ,in, Indicates that the road is clear. Indicates slow traffic on the road. Indicates that the road section is interrupted; Indicates road section At the moment The degree of capacity loss due to traffic control, with a value range of 0 to 1. A larger value indicates a more serious decline in capacity.

5. A terminal delivery planning method according to claim 4, characterized in that: The S1 also includes a linkage update mechanism, which updates the dynamic road network state matrix once at a preset time interval. , that is, when it is interrupted, the fluctuation coefficients of adjacent sections are triggered to be updated in a linked manner.

6. The terminal delivery planning method according to claim 1, characterized in that: The S2 specifically includes: S21: For each delivery node, a time sensitivity with a sudden change characteristic is constructed based on the time difference between the latest delivery time and the current system time, combined with a preset emergency response threshold. This time sensitivity is used to characterize the time urgency of the order corresponding to the delivery node, and the time sensitivity adjustment factor is used to adaptively reflect the total pressure of undelivered tasks in the current system. S22: Extract the shortest accessible path from the current vehicle location to each delivery node from the dynamic road network state matrix and calculate the actual travel time for the corresponding path. Simultaneously, a fixed radius is set around each delivery node. The number of undelivered nodes within the radius is counted and combined with their spatial distribution variance to form a spatial cluster density. Potential high-density delivery areas are identified. A weighted synthesis of path accessibility and spatial cluster density is then performed to construct a complete spatial proximity index. S23, by integrating time sensitivity and spatial proximity indicators, constructs the spatiotemporal integrated priority weight value of the delivery node. After normalization, it is divided into levels according to the dual threshold strategy to form high priority, medium priority and delayable delivery subsequences respectively, and sets sorting rules according to the spatiotemporal integrated priority weight value or spatial aggregation to construct a global priority delivery scheduling sequence.

7. A terminal delivery planning method according to claim 6, characterized in that: The S2 also includes a dynamic adjustment mechanism, which includes: if a new delivery task is detected or the road network status changes, the indicators of all nodes that have not completed the delivery are recalculated.

8. The terminal delivery planning method according to claim 1, characterized in that: The constraints of the service time window include: calculating the latest service time based on the promised delivery time of the delivery node and the current system time, and setting an adaptive buffer as the lower limit of the service time; The constraints of the energy consumption window include: constructing a unit distance energy consumption prediction model based on the current vehicle load and speed, and estimating the available energy consumption budget in combination with the remaining power to ensure that the generated path meets the dual conditions of timeliness and endurance. The unit distance energy consumption prediction model is a quadratic function of load and speed, including vehicle power coefficient, linear resistance coefficient and static energy consumption terms; the available energy consumption budget is the available energy consumption after deducting the consumed amount from the total energy and retaining the safety threshold.

9. A terminal delivery planning method according to claim 8, characterized in that: The S3 specifically includes: S31, Window Initialization: A fixed number of dispatch nodes are intercepted from the priority dispatch scheduling sequence to form a sliding window, and all arranged paths are filtered for feasibility. If a path includes a section with a disconnected connection status or does not meet the service time window and energy consumption window constraints, it is eliminated, and only the paths that meet the constraints are retained and enter the optimization process; S32, Path Optimization: A multi-objective evaluation is used to score feasible paths. The objective function simultaneously considers the total path time, cumulative energy consumption, and node spatial clustering density, and performs weighted fusion through adjustable weight coefficients. Among them, the time constraint is a rigid constraint, and the energy consumption allows a relaxation ratio to be set within the budget. The ε-constraint method is used to search for non-dominated solution sets, and the comprehensive optimal path is selected based on the objective function value.

10. A terminal delivery planning system, used to implement a terminal delivery planning method according to any one of claims 1 to 9, characterized in that: Includes the following modules: Dynamic road network data acquisition module, used to obtain basic traffic data, traffic control data and customer time efficiency demand data of each road section in the target area, and build a dynamic road network status matrix; A dispatch node weight calculation module, configured to calculate the spatiotemporal integrated priority weight value of each dispatch node based on the dynamic road network state matrix, and generate a priority dispatch scheduling sequence; The path optimization module is used to generate an optimal delivery path based on the priority delivery scheduling sequence by adopting a dynamic window algorithm based on the dual constraints of service time window and energy consumption window.

Images