Intelligent logistics scheduling system
By optimizing logistics scheduling through three-dimensional semantic density field, fluid dynamics and electromagnetic field models, intelligent vehicle loading and route selection are realized, improving the efficiency and safety of logistics scheduling and solving the inefficiency and safety hazards existing in the current technology.
Patent Information
- Application Number
- CN202510822445.6
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2025-06-19
- Publication Date
- 2025-11-11
AI Technical Summary
Existing logistics dispatch systems suffer from low vehicle volume/load utilization, disconnect between vehicle selection and route planning, high empty-running rates, serious fuel waste, and slow response to sudden orders or traffic changes, posing safety hazards and making it difficult to achieve dynamic optimization across the entire supply chain.
A three-dimensional semantic density field is used to predict cargo density, combined with a fluid dynamics model for vehicle loading, an electromagnetic field model is used to optimize the layout, and a ripple diffusion method is used to select routes, thus constructing an interdisciplinary collaborative intelligent scheduling system.
It improves space utilization, reduces overall costs, decreases damage rates, and achieves intelligent scheduling across the entire chain, solving the problems of inefficiency and safety hazards in traditional scheduling.
Smart Images
Figure CN120930969A_ABST
Abstract
Description
Technical Field
[0001] This invention relates to the field of logistics warehousing and scheduling technology, and specifically to an intelligent logistics scheduling system. Background Technology
[0002] As a core component of supply chain management, the efficiency of modern logistics scheduling systems directly impacts the cost and timeliness of social material circulation. Early logistics scheduling relied on manual experience-based decision-making; dispatchers recorded cargo parameters using paper documents and manually calculated vehicle loading plans.
[0003] Human experience-based decision-making relies on empirical estimations, resulting in an unreasonable combination of light and heavy cargo, with vehicle volume / load utilization rates of only 60% to 75%. The selection of vehicle models is disconnected from route planning, leading to high empty-running rates and significant fuel waste. When sudden orders or traffic changes occur, manual rescheduling takes a long time. Issues such as cargo stacking stability and center of gravity shift are easily overlooked.
[0004] With the introduction of computer-aided design (CAD) technology, the industry began to adopt two-dimensional bin packing algorithms (such as the BL algorithm) for spatial planning. Furthermore, intelligent algorithms such as genetic algorithms (GA) and particle swarm optimization (PSO) have also been applied to vehicle routing problems. However, these techniques have consistently failed to overcome the inherent limitations of geometric constraints and discrete optimization.
[0005] The existing technological system suffers from fundamental flaws in model building, algorithm efficiency, and interdisciplinary integration, leading the industry into a vicious cycle of high costs, low efficiency, and weak security. There is an urgent need for innovation in fundamental theories to build a new generation of intelligent scheduling systems capable of simultaneously handling multi-physics coupling and achieving dynamic optimization across the entire process. Summary of the Invention
[0006] To address the aforementioned technical problems, this invention provides an intelligent logistics scheduling system.
[0007] To solve the above-mentioned technical problems, the present invention adopts the following technical solution:
[0008] An intelligent logistics scheduling system includes:
[0009] Density prediction module: Constructs a three-dimensional semantic density field and predicts the density of unknown goods by superimposing field strengths;
[0010] Cargo loading module: Treats the vehicle's interior space as a fluid container, simulates cargo as fluids of different viscosities according to density, and dynamically obtains vehicle loading schemes through a fluid dynamics model;
[0011] Layout optimization module: The cargo is virtualized as an object with electromagnetic poles. The electromagnetic field model is used to avoid collisions by repulsion of like poles and to achieve close contact by attraction of opposite poles. This allows for the optimization of vehicle loading schemes and the calculation of loading rate.
[0012] Route selection module: Taking the vehicle as the ripple center and cost fluctuation as the ripple energy, it selects the vehicle type and route based on the loading rate and simulates the ripple diffusion path.
[0013] In one embodiment, the implementation process of the density prediction module specifically includes:
[0014] Construct a three-dimensional semantic density field coordinate system, where the X-axis represents the category level, the Y-axis represents the material density, and the Z-axis represents the packaging attenuation factor;
[0015] The BM25 algorithm is used to extract keywords for the product name. Each keyword activates the corresponding region of the three-dimensional semantic density field. The field strength value of the keyword activation is the product of the keyword weight and the industry correction coefficient.
[0016] Predicting the density ρ of the cargo:
[0017]
[0018] x m γ is the field strength value activated by the m-th keyword in the product name, M is the total number of keywords in the product name, γ is the packaging attenuation factor, α is the safety factor, and ρ0 represents the material density of the product.
[0019] In one embodiment, the weight of the keyword is calculated using the term frequency-inverse document frequency technique.
[0020] In one embodiment, the implementation process of the cargo loading module specifically includes:
[0021] Based on the density of the goods and a set threshold, the goods are divided into high-density goods, medium-density goods and low-density goods, corresponding to the sedimentation priority of the fluid, and the fluid sedimentation is simulated; among them, a greedy strategy is used to arrange the high-density goods at the bottom of the vehicle.
[0022] Each item of cargo is assigned a viscosity value σ based on its fragility parameter: σ = ρ × (1 + β); ρ is the density of the cargo, β is the fragility rating of the cargo, β = 1 when the fragility parameter of the cargo is greater than the set threshold, and β = 0 when the fragility parameter of the cargo is less than or equal to the set threshold; cargo with viscosity higher than the threshold is separated into an independent area with a buffer space reserved around it.
[0023] After stacking high-density cargo and cargo with β=1, the pressure of the remaining space in the vehicle is calculated based on the vehicle's remaining load and remaining space; medium- and low-density cargo and low-density cargo are filled along the pressure gradient direction.
[0024] In one embodiment, the calculation of the pressure of the remaining vehicle space based on the remaining vehicle load and remaining space specifically includes:
[0025]
[0026] Pressure is the force applied to the vehicle, W is the maximum load capacity of the vehicle, V is the maximum space of the vehicle, W′ is the remaining load capacity of the vehicle, and V′ is the remaining space of the vehicle.
[0027] In one embodiment, the implementation process of the layout optimization module specifically includes:
[0028] The bottom surface of the cargo is set as the S pole, the top surface of the cargo and the sides of the fragile cargo are set as the N pole, and the strength of the magnetic poles is set according to the weight of the cargo.
[0029] Calculate the interaction forces between the nearest adjacent surfaces of adjacent goods, and calculate the total field strength;
[0030] The layout of goods is iterated according to the following process until a set termination condition is reached: A single item is randomly selected for translation or rotation. If the total electric field strength of the new layout increases, the new layout is retained; if the total electric field strength of the new layout remains the same or decreases, the layout is then determined by probability. Accept the new layout, where ΔScore is the total field strength difference between the new and old layouts; T is the temperature parameter, which decreases with the number of iterations.
[0031] In one embodiment, the strength of different magnetic poles is set according to the weight of the goods, specifically including:
[0032] The strength of the S pole is 1.2 times the mass of the cargo, and the strength of the N pole is 0.8 times the mass of the cargo; the interaction force between the nearest adjacent surfaces of adjacent cargo. P1 and P2 are the magnetic pole intensities corresponding to the nearest adjacent surfaces of two adjacent goods, respectively; d is the distance between the nearest adjacent surfaces of two adjacent goods, d>0;
[0033] The interaction force F between the nearest adjacent surfaces of adjacent goods is divided into the attractive forces F generated by different magnetic poles. attract and the repulsive force F generated by the same magnetic poles repel ;
[0034] Total field strength Score = ∑F attract -∑F repel .
[0035] In one embodiment, the route selection module, which selects vehicle type and route based on the loading rate and simulated ripple diffusion path, with the vehicle as the ripple center and cost fluctuation as the ripple energy, specifically includes:
[0036] The initial energy of each vehicle is defined based on a fixed scheduling cost C and a load penalty term. Dis is the actual distance of the target route, T is the actual time the vehicle travels on the target route, W″ is the vehicle's load capacity gap, W is the vehicle's maximum load capacity, and δ is the loading rate.
[0037] The damping coefficient γ, time cost coefficient β, and distance cost coefficient α are updated based on real-time traffic and weather conditions.
[0038] For each vehicle model, an independent ripple wave is generated. Using the push-wavefront algorithm, the path branch with the least energy decay is expanded first. The energy decay formula for each vehicle model is:
[0039] E current =E initial ×e -k×f ;
[0040] f is the path complexity, and k is the vehicle efficiency coefficient;
[0041] Choose the vehicle type and route corresponding to the node where the ripple waves of different vehicle types intersect.
[0042] Compared with the prior art, the beneficial technical effects of the present invention are:
[0043] This invention solves the problems of low loading rate, uncontrollable cost, delayed response and safety hazards in traditional logistics scheduling by using fluid dynamics layered filling, electromagnetic topology collision avoidance bonding, semantic field density prediction and ripple cost diffusion. It achieves improved space utilization, reduced overall cost and reduced damage rate, and constructs a cross-disciplinary collaborative full-link intelligent scheduling paradigm. Attached Figure Description
[0044] Figure 1 This is a system framework diagram in an embodiment of the present invention. Detailed Implementation
[0045] A preferred embodiment of the present invention will now be described in detail with reference to the accompanying drawings.
[0046] like Figure 1 As shown, the present invention provides an intelligent logistics scheduling system, comprising:
[0047] Density prediction module: Constructs a three-dimensional semantic density field and predicts the density of unknown goods by superimposing field strengths;
[0048] Cargo loading module: Treats the vehicle's interior space as a fluid container, simulates cargo as fluids of different viscosities according to density, and dynamically obtains vehicle loading schemes through a fluid dynamics model;
[0049] Layout optimization module: The cargo is virtualized as an object with electromagnetic poles. The electromagnetic field model is used to avoid collisions by repulsion of like poles and to achieve close contact by attraction of opposite poles. This allows for the optimization of vehicle loading schemes and the calculation of loading rate.
[0050] Route selection module: Taking the vehicle as the ripple center and cost fluctuation as the ripple energy, it selects the vehicle type and route based on the loading rate and simulates the ripple diffusion path.
[0051] This invention introduces physical, electromagnetic, and fluid dynamics models into the logistics field, breaking away from the traditional discrete mathematics framework. By reconstructing the core logic of logistics scheduling through interdisciplinary theory, it achieves a paradigm shift in logistics scheduling from "geometric optimization" to "physical simulation" for the first time, realizing a leapfrog innovation from "experience-driven" to "physical simulation-driven".
[0052] This intelligent logistics scheduling system comprises four core modules: density prediction module, cargo loading module, layout optimization module, and route selection module. These modules collaborate through data flow.
[0053] Input: Cargo list (name, weight, volume, fragility marking), vehicle type pool data, real-time traffic conditions.
[0054] Output: Vehicle loading scheme, loading rate, and optimal route.
[0055] The data flow is as follows: density prediction → cargo loading → layout optimization → route selection; among which, the density prediction results are input into the cargo loading module, and the loading rate is fed back to the route selection module.
[0056] In one embodiment, the implementation process of the density prediction module specifically includes:
[0057] Construct a three-dimensional semantic density field coordinate system, where the X-axis represents the category level, the Y-axis represents the material density, and the Z-axis represents the packaging attenuation factor;
[0058] The BM25 algorithm is used to extract keywords for the product name. Each keyword activates the corresponding region of the three-dimensional semantic density field. The field strength value of the keyword activation is the product of the keyword weight and the industry correction coefficient.
[0059] Predicting the density ρ of the cargo:
[0060]
[0061] x m γ is the field strength value activated by the m-th keyword in the product name, M is the total number of keywords in the product name, γ is the packaging attenuation factor, α is the safety factor, and ρ0 represents the material density of the product.
[0062] In one embodiment, the weight of the keyword is calculated using the term frequency-inverse document frequency technique.
[0063] Coordinate systems can be defined in the following way:
[0064] X-axis (category level): Values are assigned according to the depth of the product category tree, for example: Electronic products = 1.0, Mobile phones = 2.3, Flagship models = 3.5.
[0065] Y-axis (material density): Preset material baseline value (metal = 8.0 g / cm³) 3 Plastic = 1.2g / cm³ 3 ).
[0066] Z-axis (Packaging Attenuation Factor): Packaging type coefficient (naked packaging = 1.0, cardboard box = 0.9, wooden box = 0.7).
[0067] Keyword extraction can be done in the following ways:
[0068] Decompose cargo names using the improved BM25 algorithm:
[0069] Input: "antistatic polyethylene granules", output keywords: ["antistatic", "polyethylene", "granules"].
[0070] The electric field strength can be calculated using the following method:
[0071] Keyword activation field strength value = Keyword weight × Industry adjustment coefficient.
[0072] Predicting the density ρ of the cargo:
[0073]
[0074] x m This is the field strength value activated by the m-th keyword in the product name, where M is the total number of keywords in the product name, γ is the packaging attenuation factor, and α is the safety factor. The safety factor is 1.2 by default (increase to 1.5 for fragile items).
[0075] In one embodiment, the process of predicting the density of goods named "thickened shockproof foam board" specifically includes:
[0076] Keywords: ["thickened", "shockproof", "foam"] → The sum of the activated field strength values = 2.1;
[0077] Material density (foam): 0.05 g / cm³ 3 →ρ0=0.05;
[0078] Packaging attenuation (cardboard box): γ = 0.9;
[0079] Predicted density: ρ=(0.05×2.1) / (0.9×1.2)≈0.097g / cm³ 3 .
[0080] Anomaly handling mechanisms include: initiating image back-calculation when the predicted density deviates from the reported value by more than 30%.
[0081] Use the YOLO model to detect cargo dimensions and calculate the visual volume V. 视觉 ;
[0082] Corrected density: ρ 修正 =Declaration quality / V 视觉 ;
[0083] Update semantic field coordinates: Optimize material density values using gradient descent.
[0084] This invention maps text to a physical parameter space, solving the problem of polysemy; it can also achieve cross-modal calibration through image detection, reducing manual intervention.
[0085] In one embodiment, the implementation process of the cargo loading module specifically includes:
[0086] Based on the density of the goods and a set threshold, the goods are divided into high-density goods, medium-density goods and low-density goods, corresponding to the sedimentation priority of the fluid, and the fluid sedimentation is simulated; among them, a greedy strategy is used to arrange the high-density goods at the bottom of the vehicle.
[0087] Each item of cargo is assigned a viscosity value σ based on its fragility parameter: σ = ρ × (1 + β); ρ is the density of the cargo, β is the fragility rating of the cargo, β = 1 when the fragility parameter of the cargo is greater than the set threshold, and β = 0 when the fragility parameter of the cargo is less than or equal to the set threshold; cargo with viscosity higher than the threshold is separated into an independent area with a buffer space reserved around it.
[0088] After stacking high-density cargo and cargo with β=1, the pressure of the remaining space in the vehicle is calculated based on the vehicle's remaining load and remaining space; medium- and low-density cargo and low-density cargo are filled along the pressure gradient direction.
[0089] This invention combines fluid mechanics with logistics loading to solve the rigid constraint problem of traditional packing algorithms; and achieves a dynamic balance between cargo protection and space utilization through viscosity coefficient.
[0090] Specifically, density classification can be done in the following way:
[0091] High density (>5g / cm³) 3 Metal products (preferably sink to the bottom);
[0092] Medium density (1-5 g / cm³) 3 Cardboard boxes;
[0093] Low density (<1g / cm³) 3 ): Foam, textiles.
[0094] Cargo is arranged in descending order of density, with high-density cargo placed preferentially at the bottom of the vehicle to simulate fluid settling. A greedy strategy is used to tightly pack high-density cargo, ensuring that the total mass does not exceed the vehicle's load limit.
[0095] In one embodiment, the calculation of the pressure of the remaining vehicle space based on the remaining vehicle load and remaining space specifically includes:
[0096]
[0097] Pressure is the force applied to the vehicle, W is the maximum load capacity of the vehicle, V is the maximum space of the vehicle, W′ is the remaining load capacity of the vehicle, and V′ is the remaining space of the vehicle.
[0098] In one embodiment, the implementation process of the layout optimization module specifically includes:
[0099] The bottom surface of the cargo is set as the S pole, the top surface of the cargo and the sides of the fragile cargo are set as the N pole, and the strength of the magnetic poles is set according to the weight of the cargo.
[0100] Calculate the interaction forces between the nearest adjacent surfaces of adjacent goods, and calculate the total field strength;
[0101] The layout of goods is iterated according to the following process until a set termination condition is reached: A single item is randomly selected for translation or rotation. If the total electric field strength of the new layout increases, the new layout is retained; if the total electric field strength of the new layout remains the same or decreases, the layout is then determined by probability. Accept the new layout, where ΔScore is the total field strength difference between the new and old layouts; T is the temperature parameter, which decreases with the number of iterations.
[0102] This invention achieves automatic obstacle avoidance and tight fit through electromagnetic polarity, replacing traditional coordinate calculations. The all-N pole design of fragile items achieves physical isolation, eliminating the need for manual marking of rules.
[0103] This invention employs Monte Carlo random perturbation: each time, one item is randomly rotated or moved, the change in field strength is calculated, and the layout optimized for field strength is preserved. The termination condition is: no improvement in field strength after 100 consecutive perturbations.
[0104] In one embodiment, the strength of different magnetic poles is set according to the weight of the goods, specifically including:
[0105] The strength of the S pole is 1.2 times the mass of the cargo, and the strength of the N pole is 0.8 times the mass of the cargo; the interaction force between the nearest adjacent surfaces of adjacent cargo. P1 and P2 are the magnetic pole intensities corresponding to the nearest adjacent surfaces of two adjacent goods, respectively; d is the distance between the nearest adjacent surfaces of two adjacent goods, d>0;
[0106] The interaction force F between the nearest adjacent surfaces of adjacent goods is divided into the attractive forces F generated by different magnetic poles. attract and the repulsive force F generated by the same magnetic poles repel ;
[0107] Total field strength Score = ∑F attract -∑F repel .
[0108] In one embodiment, the route selection module, which selects vehicle type and route based on the loading rate and simulated ripple diffusion path, with the vehicle as the ripple center and cost fluctuation as the ripple energy, specifically includes:
[0109] The initial energy of each vehicle is defined based on a fixed scheduling cost C and a load penalty term. Dis is the actual distance of the target route, T is the actual time the vehicle travels on the target route, W′′ is the vehicle's load capacity gap, W is the vehicle's maximum load capacity, and δ is the loading rate.
[0110] The load factor is calculated as follows:
[0111] Volume loading rate is the ratio of the total volume of the cargo to the total volume of the vehicle; mass loading rate is the ratio of the total mass of the cargo to the total load capacity of the vehicle.
[0112] After optimizing the vehicle loading scheme, the layout optimization module can calculate the volume loading rate and mass loading rate based on the optimized vehicle loading scheme.
[0113] The final load factor is the smaller of the volume load factor and the mass load factor.
[0114] The damping coefficient γ, time cost coefficient β, and distance cost coefficient α are updated based on real-time traffic and weather conditions.
[0115] For each vehicle model, an independent ripple wave is generated. Using the push-wavefront algorithm, the path branch with the least energy decay is expanded first. The energy decay formula for each vehicle model is:
[0116] E current =E initial ×e -k×f ;
[0117] f is the path complexity, and k is the vehicle efficiency coefficient. The vehicle efficiency coefficient k is related to the loading rate, k = k0 × (1 + (0.85 - δ)); k0 is the initial efficiency coefficient. The above formula decays faster when the loading rate is less than 85%, which can make the path of low loading rate vehicle models be eliminated more quickly in the diffusion process.
[0118] Choose the vehicle type and route corresponding to the node where the ripple waves of different vehicle types intersect.
[0119] Using the advancing wavefront algorithm, it is possible to avoid traversing all paths and naturally converge to an efficient route through energy decay.
[0120] By updating various coefficients based on real-time traffic and weather conditions, the model can respond in real time to changes in the external environment, improving its adaptability. For example, during congestion, the time cost coefficient β increases by 20%; on rainy days, the distance cost coefficient α increases by 10%.
[0121] The technical features of the above embodiments can be combined in any way. For the sake of brevity, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.
[0122] It will be apparent to those skilled in the art that the present invention is not limited to the details of the exemplary embodiments described above, and that the invention can be implemented in other specific forms without departing from its spirit or essential characteristics. Therefore, the embodiments should be considered in all respects as exemplary and non-limiting, and the scope of the invention is defined by the appended claims rather than the foregoing description. Thus, all variations falling within the meaning and scope of equivalents of the claims are intended to be included within the present invention, and no reference numerals in the claims should be construed as limiting the scope of the claims.
[0123] Furthermore, it should be understood that although this specification describes embodiments, not every embodiment contains only one independent technical solution. This narrative style is merely for clarity. Those skilled in the art should consider the specification as a whole, and the technical solutions in each embodiment can also be appropriately combined to form other embodiments that can be understood by those skilled in the art.
Claims
1. An intelligent logistics scheduling system, characterized in that, include: Density prediction module: Constructs a three-dimensional semantic density field and predicts the density of unknown goods by superimposing field strengths; Cargo loading module: Treats the vehicle's interior space as a fluid container, simulates cargo as fluids of different viscosities according to density, and dynamically obtains vehicle loading schemes through a fluid dynamics model; Layout optimization module: The cargo is virtualized as an object with electromagnetic poles. The electromagnetic field model is used to avoid collisions by repulsion of like poles and to achieve close contact by attraction of opposite poles. This allows for the optimization of vehicle loading schemes and the calculation of loading rate. Route selection module: Taking the vehicle as the ripple center and cost fluctuation as the ripple energy, it selects the vehicle type and route based on the loading rate and simulates the ripple diffusion path.
2. The intelligent logistics scheduling system according to claim 1, characterized in that, The implementation process of the density prediction module specifically includes: Construct a three-dimensional semantic density field coordinate system, where the X-axis represents the category level, the Y-axis represents the material density, and the Z-axis represents the packaging attenuation factor; The BM25 algorithm is used to extract keywords for the product name. Each keyword activates the corresponding region of the three-dimensional semantic density field. The field strength value of the keyword activation is the product of the keyword weight and the industry correction coefficient. Predicting the density ρ of the cargo: x m γ is the field strength value activated by the m-th keyword in the product name, M is the total number of keywords in the product name, γ is the packaging attenuation factor, α is the safety factor, and ρ0 represents the material density of the product.
3. The intelligent logistics scheduling system according to claim 2, characterized in that, The weight of the keywords is calculated using the term frequency-inverse document frequency technique.
4. The intelligent logistics scheduling system according to claim 1, characterized in that, The implementation process of the cargo loading module specifically includes: Based on the density of the goods and a set threshold, the goods are divided into high-density goods, medium-density goods and low-density goods, corresponding to the sedimentation priority of the fluid, and the fluid sedimentation is simulated; among them, a greedy strategy is used to arrange the high-density goods at the bottom of the vehicle. Each item of cargo is assigned a viscosity value σ based on its fragility parameter: σ = ρ × (1 + β); ρ is the density of the cargo, β is the fragility rating of the cargo, β = 1 when the fragility parameter of the cargo is greater than the set threshold, and β = 0 when the fragility parameter of the cargo is less than or equal to the set threshold; cargo with viscosity higher than the threshold is separated into an independent area with a buffer space reserved around it. After stacking high-density cargo and cargo with β=1, the pressure of the remaining space in the vehicle is calculated based on the vehicle's remaining load and remaining space; medium- and low-density cargo and low-density cargo are filled along the pressure gradient direction.
5. The intelligent logistics scheduling system according to claim 4, characterized in that, The calculation of the pressure of the vehicle's remaining space based on the vehicle's remaining load and remaining space specifically includes: Pressure refers to the force applied to the vehicle, W represents the maximum load capacity, and V represents the maximum interior space. ′ V represents the vehicle's remaining load capacity. ′ This refers to the remaining space in the vehicle.
6. The intelligent logistics scheduling system according to claim 1, characterized in that, The implementation process of the layout optimization module specifically includes: The bottom surface of the cargo is set as the S pole, the top surface of the cargo and the sides of the fragile cargo are set as the N pole, and the strength of the magnetic poles is set according to the weight of the cargo. Calculate the interaction forces between the nearest adjacent surfaces of adjacent goods, and calculate the total field strength; The layout of goods is iterated according to the following process until a set termination condition is reached: A single item is randomly selected for translation or rotation. If the total electric field strength of the new layout increases, the new layout is retained; if the total electric field strength of the new layout remains the same or decreases, the layout is then determined by probability. Accept the new layout, where ΔScore is the total field strength difference between the new and old layouts; T is the temperature parameter, which decreases with the number of iterations.
7. The intelligent logistics scheduling system according to claim 6, characterized in that, The strength of the magnetic poles is set according to the weight of the goods, specifically including: The strength of the S pole is 1.2 times the mass of the cargo, and the strength of the N pole is 0.8 times the mass of the cargo; the interaction force between the nearest adjacent surfaces of adjacent cargo. P1 and P2 are the magnetic pole intensities corresponding to the nearest adjacent surfaces of two adjacent goods, respectively; d is the distance between the nearest adjacent surfaces of two adjacent goods, d>0; The interaction force F between the nearest adjacent surfaces of adjacent goods is divided into the attractive forces F generated by different magnetic poles. attract and the repulsive force F generated by the same magnetic poles repel ; Total field strength Score = ∑F attract -∑F repel .
8. The intelligent logistics scheduling system according to claim 1, characterized in that, In the route selection module, the selection of vehicle type and route based on the loading rate and by simulating the ripple diffusion path, with the vehicle as the ripple center and cost fluctuation as the ripple energy, specifically includes: The initial energy of each vehicle is defined based on a fixed scheduling cost C and a load penalty term. Dis is the actual distance of the target route, T is the actual time the vehicle travels on the target route, W″ is the vehicle's load capacity gap, W is the vehicle's maximum load capacity, and δ is the loading rate. The damping coefficient γ, time cost coefficient β, and distance cost coefficient α are updated based on real-time traffic and weather conditions. For each vehicle model, an independent ripple wave is generated. Using the push-wavefront algorithm, the path branch with the least energy decay is expanded first. The energy decay formula for each vehicle model is: AND current =And initial ×e -k×f ; f is the path complexity, and k is the vehicle efficiency coefficient; Choose the vehicle type and route corresponding to the node where the ripple waves of different vehicle types intersect.