Logistics distribution path dynamic planning method based on multi-objective optimization

Abstract

The invention relates to the technical field of logistics automation and intelligent scheduling, and discloses a multi-objective optimization logistics distribution path dynamic planning method, which comprises the following steps: S1, mapping a logistics distribution area and vehicles, clients and constraint entities in the area to a continuous coordinate space; s2, in the continuous coordinate space, optimizing cost, time, priority and constraints based on the mapped clients and constraint entities, and respectively constructing corresponding decision force fields; s3, based on a preset strategy adjustment matrix, carrying out linear transformation and combination on the plurality of constructed decision force fields so as to synthesize total decision force acting on the space; and S4, taking the synthesized total decision-making force as an external driving force. According to the method, through the strategy adjustment matrix and the path dynamic evolution equation, the high-level strategy and the path physical characteristics are deeply coupled, and refined, smooth and efficient regulation and control of multi-target dynamic path planning are realized.

Description

Technical Field

[0001] This invention relates to the field of logistics automation and intelligent scheduling technology, specifically to a dynamic planning method for multi-objective optimization of logistics distribution routes. Background Technology

[0002] In existing technological practices, logistics delivery route planning problems are usually modeled as vehicle routing problems (VRP) or their variants. The mainstream approach to solving such problems relies on combinatorial optimization algorithms, such as genetic algorithms, ant colony algorithms, or simulated annealing algorithms, to find the optimal or suboptimal delivery order and route under a series of constraints. When faced with multiple optimization objectives such as cost, time, and service priority, a common approach is the weighted summation method, which pre-determines a fixed weight coefficient for each optimization objective, thereby transforming the multi-objective problem into a single-objective optimization problem for solution. For dynamically changing environments, such as the emergence of new orders or changes in traffic conditions, the strategy typically adopted by existing systems is to trigger a complete recalculation to generate new route solutions.

[0003] While the aforementioned technical solutions address the path planning problem in logistics and distribution to some extent, they still have shortcomings in dealing with the increasing complexity and dynamic demands of modern logistics. Firstly, the linear weighted summation of multiple optimization objectives relies on the assumption that these objectives are independent. This is significantly inconsistent with complex decision-making scenarios in the real world. In actual operations, the trade-off between objectives such as cost and time is often dynamic and non-linear. For example, for an urgent order that is about to exceed its time limit, the weight of the time objective should dynamically and overwhelmingly outweigh the cost objective. The fixed weighting coefficients in existing technologies cannot capture this complex, context-dependent strategic coupling, preventing the implementation of truly refined and scenario-based scheduling strategies.

[0004] Secondly, regarding the response mechanism for dynamic events, the existing technology generally adopts the full replanning approach, which has huge computational overhead and significant response delay. Since the vehicle routing problem is essentially an NP-hard combinatorial optimization problem, a global solution is performed from scratch every time the environment changes, which often consumes a lot of valuable computation time. This delay makes the system decision always lag behind the actual changes, and there is a lack of smooth transition between the newly generated path and the old path, which often manifests as discontinuous changes. This causes trouble for the driver's actual execution and damages the overall efficiency and stability of the path.

[0005] Furthermore, existing technologies typically separate "planning the optimal path" from "the dynamic characteristics of the path" during the optimization process. The output path scheme is essentially a series of static target points, while the physical properties of the path itself, such as smoothness and curvature changes, are not directly related to the high-level business strategy in the core optimization logic. For example, the system cannot directly generate a path that differs in dynamic characteristics (such as planning inertia) based on the strategy instruction of "conservative stability" or "aggressive speed," thus causing a disconnect between the strategy intent and the execution behavior. Summary of the Invention

[0006] To address the shortcomings of existing technologies, this invention provides a multi-objective optimization dynamic planning method for logistics distribution routes. This method solves the problems of existing logistics route planning methods, such as difficulty in describing the complex strategic coupling relationships between multiple objectives, unsmooth and inefficient response to dynamic events, and the disconnect between high-level scheduling strategies and the dynamic characteristics of the final route.

[0007] To achieve the above objectives, the present invention provides the following technical solution: a multi-objective optimization dynamic planning method for logistics distribution routes, comprising the following steps:

[0008] S1. Map the logistics and distribution area, as well as the vehicles, customers, and constrained entities within that area, to a continuous coordinate space.

[0009] S2. In this continuous coordinate space, based on the mapped customer and constraint entities, and optimizing for cost, time, priority and constraints, construct corresponding decision force fields respectively;

[0010] S3. Based on the preset strategy adjustment matrix, perform linear transformation and combination on the constructed multiple decision force fields to synthesize the total decision force acting on the space;

[0011] S4. Using the synthesized total decision-making force as an external driving force, and based on the current position of the mapped vehicle as the initial condition, solve the preset path dynamic evolution equation to generate the planned path of the vehicle.

[0012] Preferably, in step S1, the step of mapping the entity to a continuous coordinate space includes:

[0013] The vehicle's location information is obtained in real time through the vehicle-mounted positioning terminal. The location, time window, and business priority information related to the customer are retrieved from the enterprise resource planning system. Real-time traffic congestion or road control information corresponding to the constrained entity is obtained by calling a third-party map service interface.

[0014] The vehicle, customer, and constraint information obtained above are mapped together to the continuous coordinate space.

[0015] Preferably, in step S2, the step of constructing the corresponding decision force field includes:

[0016] Based on the location of the mapped customer, a cost gravity field is constructed to characterize the driving mileage target.

[0017] Based on the client's time window and current time as mapped, a time pressure field is constructed in which the magnitude of the field force dynamically increases over time.

[0018] Based on the business priorities of the customers mapped, a priority charge field is constructed to characterize the importance of the customers;

[0019] Based on the constraint entities of the mapping, a constraint repulsion field is constructed to avoid negative regions.

[0020] Preferably, in the construction process of the time-pressure field, the dynamic enhancement of its force magnitude is achieved through a time-weighting function, the formula of which is:

[0021]

[0022] In the formula: t is the current physical time; The delivery deadline for customer i; α is the preset time sensitivity coefficient; w t This is the time weighting function.

[0023] Preferably, in step S3, the step of synthesizing the total decision-making power further includes:

[0024] The constructed multiple decision force fields are mathematically represented as a single input force vector bundle;

[0025] The input force vector bundle is multiplied by the strategy adjustment matrix to obtain the output force vector bundle modulated by the strategy.

[0026] The total decision force is obtained by summing the vector components in the output force vector bundle.

[0027] The formula for the output force vector beam is:

[0028]

[0029] In the formula: P is the input force vector bundle; P is the strategy adjustment matrix; This is for outputting a force vector beam.

[0030] Preferably, the step of synthesizing the overall decision-making power further includes:

[0031] The diagonal elements of the policy adjustment matrix are used to characterize the basic weights for a single optimization objective, while the off-diagonal elements are used to characterize the synergistic enhancement or inhibition relationships between different optimization objectives.

[0032] Preferably, in step S4, the step of generating the planned path includes:

[0033] The path dynamic evolution equation is a second-order dynamic response equation, which structurally includes a planning inertia term, a planning damping term, a path intrinsic stiffness term, and the synthesized total decision force.

[0034] Preferably, the values ​​of planning inertia and planning damping in the second-order dynamic response equation are calculated using the following formula:

[0035] m = f m (P); c = f c (P);

[0036] In the formula: m is the planning inertia; c is the planning damping; P is the strategy adjustment matrix; f m f is the mapping function from the policy adjustment matrix to the planning inertia; c Let be the mapping function from the policy adjustment matrix to the planning damping.

[0037] Preferably, a dynamic response process is also included:

[0038] The system continuously monitors dynamic events such as new orders, customer cancellations, or changes in road conditions. When such dynamic events occur, the system updates the decision force field in real time and continues to solve the path dynamic evolution equation based on the updated decision force field and the current strategy adjustment matrix, so that the planned path can smoothly evolve towards a new optimal solution.

[0039] Preferably, it also includes a strategy adjustment process:

[0040] In response to external scheduling instructions, the element values ​​in the strategy adjustment matrix are modified in real time, thereby instantly changing the synthesis method of the total decision-making force and the response characteristics of the path dynamic evolution equation.

[0041] This invention provides a dynamic programming method for multi-objective optimization of logistics distribution routes. It has the following beneficial effects:

[0042] 1. By introducing a strategy adjustment matrix, the synergistic enhancement or inhibition relationship between multiple optimization objectives is defined mathematically and in a matrix form. This avoids the limitations of simply superimposing the weights of each objective in traditional methods. It makes the formulation of scheduling strategies no longer just a matter of allocating importance, but can more profoundly depict the complex coupling relationship between different objectives in specific situations. This gives the logistics route planning system a high degree of decision-making depth and strategic flexibility, enabling more refined and scenario-based operation management.

[0043] 2. By transforming the path planning problem into a process of solving the dynamic evolution equation of the path in a continuous space, the system's response to external dynamic events becomes more efficient and smoother. When a new order or real-time road condition changes, the system only needs to update the corresponding decision force field and continue to solve the evolution equation. The planned path will then naturally transition to the new optimal solution, avoiding the delays and path abrupt changes caused by recalculation in traditional methods. The path generation method based on physical evolution ensures the continuity of planning results and high adaptability to dynamic environments.

[0044] 3. By designing the planning inertia and planning damping, which affect the dynamic response characteristics of the path, as variables derived from the strategy adjustment matrix, an intrinsic link is established between high-level business strategies and underlying path physical behavior. When the delivery strategy changes, not only will the overall decision-making force driving the path evolution be adjusted accordingly, but the smoothness and flexibility of the path itself will also change accordingly. The deep coupling of strategy and execution realizes the overall and systematic control of the path planning process, so that the final generated path can accurately reflect the preset strategic intent in both form and behavior. Attached Figure Description

[0045] Figure 1 This is a schematic diagram of the method flow of the present invention;

[0046] Figure 2 This is a schematic diagram of the architecture of a continuous coordinate space according to an embodiment of the present invention;

[0047] Figure 3 This is a schematic diagram of the overall decision-making power synthesis process according to an embodiment of the present invention;

[0048] Figure 4 This is a schematic diagram of the path generation process of the strategy adjustment matrix in an embodiment of the present invention. Detailed Implementation

[0049] The technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. 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.

[0050] Please see the appendix Figure 1 -Appendix Figure 4 This invention provides a multi-objective optimization dynamic planning method for logistics distribution routes, comprising the following steps:

[0051] S1. Map the logistics and distribution area, as well as the vehicles, customers, and constrained entities within that area, to a continuous coordinate space.

[0052] S2. In this continuous coordinate space, based on the mapped customer and constraint entities, and optimizing for cost, time, priority and constraints, construct the corresponding decision force fields respectively;

[0053] S3. Based on the preset strategy adjustment matrix, perform linear transformation and combination on the constructed multiple decision force fields to synthesize the total decision force acting on the space;

[0054] S4. Using the synthesized total decision-making force as the external driving force, and based on the mapped current position of the vehicle as the initial condition, solve the preset path dynamic evolution equation to generate the planned path of the vehicle.

[0055] In this embodiment, step S1, namely the environment mapping step, serves as the data foundation and spatial framework for all subsequent calculations and analyses in this invention. This step aims to construct a digital environment that can be used for subsequent physical modeling and dynamic evolution.

[0056] First, the system acquires information about the vehicle entity through real-time communication with the positioning module mounted on the delivery vehicle. This location information (such as longitude and latitude coordinates) serves as the initial condition for solving the subsequent path dynamic evolution equations and as a dynamic reference point for the planning entity in continuous coordinate space.

[0057] Secondly, regarding the acquisition of customer entity information, the system establishes a deep data interface with the enterprise's backend business management platform. The system can retrieve and analyze all customer data related to the current delivery batch. This data forms the core of the subsequent decision-making force field construction, specifically including:

[0058] Customer location information: This refers to the precise delivery address of each customer. The system needs to convert this from text address format to geographic coordinates within the same coordinate system as the vehicle location information, which will form the basis for subsequently constructing a cost gravitational field.

[0059] Time window attribute: This refers to the delivery time range required by the customer, such as "between 9:00 AM and 11:00 AM". This time attribute is a key input for building a time pressure field.

[0060] Business Priority Attribute: This attribute is quantified into a numerical value based on the customer's contract level (e.g., strategic partner, VIP, regular customer) or the urgency of the order (e.g., emergency medicine order, regular goods order).

[0061] Furthermore, regarding the acquisition of information on various constraint entities in the environment, the system achieves this by calling a fully functional third-party map service application programming interface (API). This allows the system to collect various external factors affecting the practical feasibility of paths into a unified model framework. These constraint entities can be categorized as follows:

[0062] Dynamic constraints: such as real-time traffic congestion information, accident locations, and traffic control information due to temporary events or weather conditions obtained through APIs.

[0063] Static constraints: such as road control information inherent in map data, including one-way streets, truck restricted areas, road height restrictions, bridge weight restrictions, long-term construction areas, etc.

[0064] Finally, after collecting the above information, the information from different sources and in different formats is normalized and mapped to a preset, continuous Cartesian coordinate space. In this space, vehicles and each customer become a point mass carrying attributes such as location, time, and priority; while various constraint entities are abstracted into specific geometric regions or boundaries with specific attributes in this space.

[0065] In this embodiment, step S2, namely the decision field construction step, builds upon the digital environment constructed in step S1. In the continuous coordinate space, a decision force field is constructed for each independent optimization objective, unifying the originally independent or even conflicting business needs (such as cost, timeliness, priority, and compliance) under the framework of the vector field, thereby providing the possibility for strategic integration and superposition in the subsequent step S3.

[0066] Specifically, for different optimization objectives, this step constructs the following decision force fields respectively:

[0067] The construction of the cost gravity field involves each customer entity, which carries precise coordinate information and is mapped in step S1. In the method of this invention, each customer point is abstracted as a gravity source, similar to a point mass in physics, generating a cost gravity field in space. At any point in the continuous coordinate space, the virtual point representing the planned path will be subject to the gravitational force from all unserved customer points. The magnitude of this gravity can be set to be inversely proportional to the square of the distance, and the direction is explicitly pointed to the gravity source (i.e., the customer point). The purpose of this design is to provide a fundamental driving force for the evolution of the path and to seek the optimal solution for the total path length on a macroscopic level.

[0068] The construction of the time pressure field relies on the time window attributes of the customer entities mapped in step S1, especially their delivery deadlines, which are closely related to the current physical time. The core idea is that when the current time is still far from the customer's required delivery deadline, the force field is weak and has little impact on path planning. As time passes and the current time approaches the delivery deadline, the strength of the force field increases sharply, forcing the path to shift towards the customer who is about to exceed the time limit. This dynamically enhanced characteristic is achieved by introducing a time weighting function, the formula of which is:

[0069]

[0070] In the formula: t is the current physical time; α is the delivery deadline for customer i, which is obtained from the business system in step S1; α is a preset time sensitivity coefficient, which can be configured by the system administrator according to business needs to control the severity of time pressure increases; w t This is a time-weighted function, representing the pressure coefficient that increases over time.

[0071] The construction of the priority charge field is based on the business priority attribute representing the importance of customers obtained in step S1. In the embodiments of the present invention, each customer entity is analogous to a particle carrying a charge. The magnitude of its "charge" is positively correlated with its business priority value. A VIP customer or urgent order that is given a high priority value will be regarded as a particle carrying a large amount of positive charge, thus generating a stronger priority "electric field" around it. When the planned path travels through this space, the magnitude of the electric field force it experiences will be proportional to the customer's priority, thereby attracting it more powerfully to the higher value or more urgent target among many customers waiting to be served.

[0072] The construction of the constraint repulsion field is based on the various constraint entities mapped in step S1, such as real-time traffic congestion areas, road construction sites, and truck restriction areas. In the model, when the planned path approaches or attempts to cross these areas, it will be subject to a strong repulsive force pointing from the inside of the area to the outside. The magnitude of this repulsive force is inversely proportional to the distance from the path point to the boundary of the area. The closer the distance, the greater the repulsive force. This design enables the path to actively and smoothly bypass these negative areas during evolution, avoiding the generation of unrealistic or illegal route schemes.

[0073] In summary, through the execution of step S2, this invention transforms a multi-objective logistics distribution problem into a physical system in the same space, in which multiple decision-making force fields, such as cost gravity field, time pressure field, priority charge field and constraint repulsion field, interact and couple with each other.

[0074] In this embodiment, step S3, namely the overall decision-making force synthesis step, is the decision-making core and strategy center of the technical solution of the present invention. This step will transform and combine the multiple decision-making force fields that may conflict with each other, which are constructed in step S2 and represent different optimization objectives, according to the preset and flexibly adjustable delivery strategy, and finally synthesize a unique overall decision-making force with a clear direction. This overall decision-making force will serve as the only external driving source for the dynamic evolution of the path in step S4.

[0075] In actual logistics and delivery scenarios, there are often complex nonlinear coupling relationships between different optimization objectives, rather than simple linear superposition. For example, excessively pursuing the lowest cost may lead to time delays, while unconditionally satisfying timeliness may significantly increase the mileage. Traditional weighted summation methods are difficult to characterize this complex decision-making logic. Therefore, this invention introduces the core concept of "strategy adjustment matrix" to solve this problem.

[0076] Specifically, the execution flow for this step is as follows:

[0077] First, the input force field is vectorized. At any point in the continuous coordinate space, the force vectors generated by the various decision force fields (cost gravitational field, time pressure field, priority charge field, and constraint repulsion field) constructed in step S2 are mathematically organized and arranged into a column vector. This column vector is defined as the input force vector bundle, which is the set of all original driving forces at that point in space.

[0078] Secondly, the core policy transformation operation is performed. This operation is achieved by multiplying a preset, user-defined policy adjustment matrix with the input force vector bundle, thereby obtaining a policy-modulated output force vector bundle. The mathematical expression for this process is:

[0079]

[0080] In the formula: The input force vector bundle has each component corresponding to the force vector generated at that point by a decision force field constructed in step S2; P is the policy adjustment matrix, which encodes the weights, cooperation, and inhibition relationships between different optimization objectives. The output force vector bundle consists of force vectors whose components have been adjusted by the strategy and whose weights and directions have been redistributed.

[0081] The design of this strategy adjustment matrix is ​​the essence of this step, and the physical meaning and business implications of its internal elements are clearly defined:

[0082] The diagonal elements of the matrix (P) ii ): Used to characterize the basic weight or importance of a single optimization objective. For example, the larger the value of the element corresponding to the cost gravity field on the diagonal, the greater the consideration of cost control in the overall strategy. This function is similar to the traditional weighting method and forms the basis of strategy adjustment.

[0083] The off-diagonal elements of the matrix (P) ij(where i ≠ j): This is an innovative design of the present invention, used to characterize the cross-coupling effect between different optimization objectives, that is, how the state change of objective j will affect the force of objective i. This makes it possible to define complex, nonlinear, conditional strategies, for example, a positive off-diagonal element P. 12 (Where row 1 corresponds to cost and column 2 corresponds to time) can represent a synergistic relationship. When the effect of the time pressure field (objective 2) is significantly enhanced, the weight of the cost attraction field (objective 1) should also be enhanced. Its physical meaning is that even when time is tight, the system should pay more attention to cost and avoid the route becoming extremely uneconomical due to excessive pursuit of speed. Conversely, a negative off-diagonal element can represent an inhibitory relationship, such as allowing for a moderate relaxation of cost restrictions when serving high-priority customers.

[0084] Finally, the vector synthesis of the total decision-making force is performed by summing all vector components in the output force vector bundle obtained after the above matrix operations, thereby obtaining the final, unique, and synthesized total decision-making force acting on this spatial point, which combines all business strategies.

[0085] In summary, step S3, by introducing and applying the strategy adjustment matrix, provides guidance for generating a feasible planning path that takes into account multiple objectives in step S4.

[0086] In this embodiment, step S4, namely the path generation step, is the final execution link of the technical solution of the present invention. The fundamental purpose of this step is to take the total decision-making power synthesized in step S3 as the only external driving source, and generate a smooth and feasible planned path that starts from the current position of the vehicle, connects various service targets, and satisfies all constraints by solving a preset path dynamic evolution equation.

[0087] Specifically, this invention employs a second-order dynamic response equation as the path dynamic evolution equation, treating the trajectory q(s) of the planned path in continuous coordinate space (where s is the path evolution parameter, which can be understood as the arc length of the path) as an elastic body with mass and damping. Its dynamic behavior is jointly determined by the following core components:

[0088] Planning inertia term: This term is mathematically related to the second derivative of the path trajectory. Relatedly, a large planning inertia will cause the path to tend to maintain its current direction, thus generating a smoother trajectory with less curvature change, which is consistent with the kinematics of vehicles in the real world and avoids unrealistic sharp turns.

[0089] Planning damping term: This term is mathematically related to the first derivative of the path trajectory. Related, used to suppress possible oscillations and ensure that the path can converge to the optimal solution stably and quickly.

[0090] Path intrinsic stiffness term: This term represents the elastic restoring force of the path itself, which is used to punish excessive stretching or bending of the path, ensuring the intrinsic integrity and geometric rationality of the path as a continuous curve.

[0091] Overall decision-making force: This is the unique overall decision-making force synthesized in step S3. It is the external force that drives the entire path evolution process and guides the path toward the overall optimal direction of all optimization objectives.

[0092] The innovation of this invention lies in the fact that the key physical parameters in the above-mentioned second-order dynamic response equation, namely the planned inertia and planned damping, are not fixed constants. Their values ​​are determined by a preset mapping function based on the strategy adjustment matrix defined in step S3, as shown in the formula:

[0093] m = f m (P); c = f c (P);

[0094] In the formula: m is the planning inertia; c is the planning damping; P is the strategy adjustment matrix; f m f is the mapping function from the policy adjustment matrix to the planning inertia; c These are the mapping functions from the policy adjustment matrix to the planning damping. The purpose of these two mapping functions is to extract and transform the complex policy information of a multidimensional matrix into scalar physical parameters that can directly affect the dynamic behavior of the path.

[0095] This design establishes a direct and dynamic link between high-level business strategies and the underlying path physical characteristics. For example, when a dispatcher adjusts the strategy adjustment matrix to respond to an emergency, making it exhibit an aggressive "speed-first" strategy, the mapping function f... m It can be designed to output a smaller planning inertia m. A smaller inertia means that the path's "turning" will become more "agile," able to respond to changes in force with a smaller turning radius. Similarly, the mapping function f m The damping c will also be adjusted accordingly to ensure that this more flexible path can quickly stabilize under the new strategy.

[0096] In solving the problem, the current position of the vehicle obtained in step S1 is used as the initial condition for the path dynamic evolution equation. The numerical integration method is used to solve the equation. The solution process will generate a series of discrete spatial coordinate points. Connecting these points will form the final planned path that integrates all optimization objectives and strategies.

[0097] This path generation method based on dynamic evolution equations supports dynamic response and strategy adjustment. When changes in the external environment (such as new orders or traffic congestion) cause an update to the overall decision-making power in step S3, this step does not need to discard the currently planned part of the path and completely recalculate. Instead, it uses the current end state of the path as the new initial condition and continues to solve the evolution equation under the new overall decision-making power, allowing the path to smoothly and seamlessly transition to the new optimal solution. Similarly, when the strategy adjustment matrix is ​​modified in real time, its impact is immediately reflected in the subsequent path generation through changes in overall decision-making power, planning inertia, and planning damping, giving the entire planning system flexibility and responsiveness.

[0098] Although embodiments of the invention have been shown and described, it will be understood by those skilled in the art that various changes, modifications, substitutions and alterations can be made to these embodiments without departing from the principles and spirit of the invention, the scope of which is defined by the appended claims and their equivalents.

Claims

1. A dynamic programming method for multi-objective optimization of logistics distribution routes, characterized in that, Includes the following steps: S1. Map the logistics and distribution area, as well as the vehicles, customers, and constrained entities within that area, to a continuous coordinate space. S2. In this continuous coordinate space, based on the mapped customer and constraint entities, and optimizing for cost, time, priority and constraints, construct corresponding decision force fields respectively; S3. Based on the preset strategy adjustment matrix, perform linear transformation and combination on the constructed multiple decision force fields to synthesize the total decision force acting on the space; S4. Using the synthesized total decision-making force as an external driving force, and based on the mapped vehicle's current position as the initial condition, solve a preset path dynamic evolution equation to generate the vehicle's planned path.

2. The multi-objective optimization dynamic programming method for logistics distribution routes according to claim 1, characterized in that, In step S1, the step of mapping the entity to a continuous coordinate space includes: The vehicle's location information is obtained in real time through the vehicle-mounted positioning terminal. The location, time window, and business priority information related to the customer are retrieved from the enterprise resource planning system. Real-time traffic congestion or road control information corresponding to the constrained entity is obtained by calling a third-party map service interface. The vehicle, customer, and constraint information obtained above are mapped together to the continuous coordinate space.

3. The multi-objective optimization dynamic programming method for logistics distribution routes according to claim 2, characterized in that, In step S2, the steps for constructing the corresponding decision force field include: Based on the location of the mapped customer, a cost gravity field is constructed to characterize the driving mileage target. Based on the client's time window and current time as mapped, a time pressure field is constructed in which the magnitude of the field force dynamically increases over time. Based on the business priorities of the customers mapped, a priority charge field is constructed to characterize the importance of the customers; Based on the constraint entities of the mapping, a constraint repulsion field is constructed to avoid negative regions.

4. The multi-objective optimization dynamic programming method for logistics distribution routes according to claim 3, characterized in that, In the construction process of the time-pressure field, the dynamic enhancement of its force magnitude is achieved through a time-weighting function, the formula of which is: In the formula: t is the current physical time; α is the delivery deadline for customer i; α is the preset time sensitivity coefficient. w t This is the time weighting function.

5. The multi-objective optimization dynamic programming method for logistics distribution routes according to claim 4, characterized in that, In step S3, the step of synthesizing the total decision-making power further includes: The constructed multiple decision force fields are mathematically represented as a single input force vector bundle; The input force vector bundle is multiplied by the strategy adjustment matrix to obtain the output force vector bundle modulated by the strategy. The total decision force is obtained by summing the vector components in the output force vector bundle. The formula for the output force vector beam is: In the formula: P is the input force vector bundle; P is the strategy adjustment matrix; This is for outputting a force vector beam.

6. The multi-objective optimization dynamic programming method for logistics distribution routes according to claim 5, characterized in that, The step of synthesizing overall decision-making power further includes: The diagonal elements of the policy adjustment matrix are used to characterize the basic weights for a single optimization objective, while the off-diagonal elements are used to characterize the synergistic enhancement or inhibition relationships between different optimization objectives.

7. The multi-objective optimization dynamic programming method for logistics distribution routes according to claim 6, characterized in that, In step S4, the step of generating the planned path includes: The path dynamic evolution equation is a second-order dynamic response equation, which structurally includes a planning inertia term, a planning damping term, a path intrinsic stiffness term, and the synthesized total decision force.

8. The multi-objective optimization dynamic programming method for logistics distribution routes according to claim 7, characterized in that, The values ​​of planned inertia and planned damping in the second-order dynamic response equation are calculated using the following formulas: m=f m (P);c=f c (P); In the formula: m is the planning inertia; c is the planning damping; P is the strategy adjustment matrix; f m f is the mapping function from the policy adjustment matrix to the planning inertia; c Let be the mapping function from the policy adjustment matrix to the planning damping.

9. The multi-objective optimization dynamic programming method for logistics distribution routes according to claim 8, characterized in that, It also includes a dynamic response process: The system continuously monitors dynamic events such as new orders, customer cancellations, or changes in road conditions. When such dynamic events occur, the system updates the decision force field in real time and continues to solve the path dynamic evolution equation based on the updated decision force field and the current strategy adjustment matrix, so that the planned path can smoothly evolve towards a new optimal solution.

10. The multi-objective optimization dynamic programming method for logistics distribution routes according to claim 9, characterized in that, It also includes the strategy adjustment process: In response to external scheduling instructions, the element values ​​in the strategy adjustment matrix are modified in real time, thereby instantly changing the synthesis method of the total decision-making force and the response characteristics of the path dynamic evolution equation.

Images