Distribution route optimization system based on cloud computing

Abstract

The invention relates to the technical field of distribution optimization, in particular to a distribution route optimization system based on cloud computing, which comprises a space node construction module, a batch fusion identification module, a stability level configuration module, a node path screening module and a cloud scheduling optimization module. According to the method, through constructing a triple constraint rule of a space direction, a service range and a time window, accurate extraction of potential association between order nodes is realized, stability grade distinguishing of delivery batches is completed in combination with a tolerance standard of agricultural product categories under multi-dimensional environment indexes, road node environment indexes are screened, and the stability of the delivery batches is improved. It is ensured that the path selection process conforms to the actual transportation requirements of agricultural products, linkage matching from path generation to batch scheduling is achieved, the uncertainty of artificial experience line setting and scheduling is avoided, the path construction rationality, batch fusion accuracy and node scheduling intelligence are improved, and the path selection efficiency is improved. And the safety of the path selection process and the timeliness guarantee of an agricultural product circulation chain are enhanced.

Description

Technical Field

[0001] This invention relates to the field of delivery optimization technology, and in particular to a cloud computing-based delivery route optimization system. Background Technology

[0002] The field of distribution optimization technology involves the rational arrangement of transportation routes, scheduling methods, and distribution resource allocation for goods and materials in the supply chain to reduce distribution costs, improve transportation efficiency, and ensure timeliness. This includes optimized design of transportation routes, coordinated scheduling of transportation resources, dynamic adjustment of transportation time, and real-time management of the transportation process by information systems. It integrates logistics management, route calculation methods, vehicle scheduling strategies, and transportation collaboration mechanisms to form a comprehensive optimization system for various distribution demand scenarios. Traditional agricultural product logistics operations refer to the process of bulk transfer of agricultural products from the production area to the consumer market or processing enterprise through transportation tools. It usually adopts the method of batch transportation, organizes transportation plans according to the harvest time and market demand, sets transportation routes based on human experience, and completes the organization and execution of the circulation process through multiple transfers, manual telephone dispatch, and paper document management.

[0003] In the current agricultural product distribution process, route planning and scheduling schemes mainly rely on manual experience, lacking systematic modeling of the spatiotemporal distribution relationships between orders. In fresh produce transportation scenarios involving high-frequency fluctuations, batch confusion is easily caused by path redundancy and transportation delays. Manual scheduling is limited by its real-time response capability and information integration efficiency, and cannot achieve real-time filtering and path reconstruction for changes in the distribution node environment. Especially in the mixed transportation of various types of agricultural products, the lack of differentiation of environmental stability requirements leads to different batches sharing path nodes, resulting in the risk of quality loss. Paper-based management also limits the traceability management and dynamic allocation of the entire distribution process, easily leading to problems such as waste of distribution resources and decreased execution efficiency. Summary of the Invention

[0004] The purpose of this invention is to address the shortcomings of existing technologies by proposing a cloud computing-based delivery route optimization system.

[0005] To achieve the above objectives, the present invention adopts the following technical solution: A cloud computing-based delivery route optimization system includes:

[0006] The spatial node construction module obtains the latitude and longitude of delivery and time node in each agricultural product order, calculates whether the cosine value of the angle between the spatial direction vectors between nodes exceeds the path direction cosine similarity threshold, calculates whether the Euclidean distance between nodes is lower than the service radius threshold, combines whether the delivery time difference is less than the time window grace period, marks the node combination that simultaneously satisfies the three constraints, and generates a set of nodes associated with direction and time distance.

[0007] Based on the set of direction and time distance related nodes, the batch fusion identification module performs batch merging and marking on order nodes, assigns a unique number to each batch, and generates a fusion delivery batch structure set.

[0008] The stability level configuration module retrieves the environmental tolerance standards of each batch of agricultural products in the dimensions of vibration, temperature and humidity based on the integrated delivery batch structure set, constructs the stability level differentiation standard, and generates a set of batch environmental stability levels.

[0009] Based on the batch environmental stability level, the node path filtering module connects to the node data in the cloud road node platform to screen whether the vibration amplitude and temperature and humidity fluctuations meet the minimum stability level, retains compliant nodes and extracts path segments with complete structure, and generates a set of schedulable path segments.

[0010] As a further embodiment of the present invention, the set of direction and time distance associated nodes includes direction vector similarity identifier, service radius compliance identifier, and time window overlap identifier; the fused delivery batch structure set includes order number and batch identifier mapping relationship, batch order quantity statistics information, and batch unique identifier code; the batch environmental stability level set includes vibration tolerance level, temperature adaptability level, and humidity control level; and the schedulable path segment set includes environmental level matching path nodes, structurally complete path segment information, and path segment connection relationship.

[0011] As a further aspect of the present invention, the spatial node construction module includes:

[0012] The node information extraction submodule extracts the delivery latitude and longitude and delivery time node based on the delivery information in each agricultural product order, parses the latitude and longitude values, checks the data integrity and format consistency, and generates a spatiotemporal information group of delivery nodes.

[0013] The spatial relationship determination submodule calculates the cosine of the angle between the spatial direction vectors based on the latitude and longitude coordinates of each pair of nodes in the delivery node spatiotemporal information group, filters node combinations with cosine values ​​not lower than the path direction cosine similarity threshold, calculates the Euclidean distance of the corresponding node pair, filters node combinations with Euclidean distance not higher than the service radius threshold, and cross-compares to generate directional distance matching node pairs.

[0014] The direction and time distance matching submodule calculates the delivery time difference between nodes based on the delivery time of each pair of nodes in the direction and distance matching node pair, judges it against the time window grace period, filters the node combination that meets the time difference condition, and obtains the direction and time distance associated node set.

[0015] As a further aspect of the present invention, the batch fusion identification module includes:

[0016] The node batch merging submodule extracts the order number information and associated node combination structure within the set of nodes associated with direction and time distance based on the aforementioned set. It constructs a mutual belonging relationship graph for node combinations with cross-pointing relationships and establishes an order merging identifier set by grouping any node in the graph structure with all connected nodes in the same group.

[0017] The batch order counting submodule counts the number of order nodes belonging to each batch based on the order merging identifier set, extracts the corresponding order number set under each batch identifier, performs a length detection operation, records the number of orders, and constructs a batch order distribution array based on the quantity distribution to generate batch order quantity distribution data.

[0018] The batch number generation submodule assigns a unique number to each batch based on the batch order quantity distribution data, calculates and obtains the merged batch identifier value, binds the merged batch identifier value with the batch number, and establishes a merged delivery batch structure set.

[0019] As a further aspect of the present invention, the stability level configuration module includes:

[0020] The category environmental standard extraction submodule retrieves the corresponding agricultural product categories in each batch according to the integrated delivery batch structure set, obtains the environmental tolerance standard values ​​corresponding to the category under the three dimensions of vibration, temperature and humidity, performs matching and retrieval according to the category number and batch index, groups and stores them by batch number, and generates a batch category environmental standard dataset.

[0021] The standard intersection acquisition submodule calculates and obtains the batch environment intersection fitness value based on the batch category environmental standard dataset and the three standard values ​​of vibration, temperature and humidity corresponding to different categories in each batch.

[0022] The hierarchical structure generation submodule combines the set environmental stability level division benchmark, matches and judges the batch environmental intersection fitness value with the level threshold, judges the segment of each batch intersection value according to the threshold range and assigns a level label, and generates a batch environmental stability level set.

[0023] As a further aspect of the present invention, the node path filtering module includes:

[0024] The node environment filtering submodule, based on the minimum vibration tolerance standard value, minimum temperature tolerance standard value, and minimum humidity tolerance standard value corresponding to each batch in the batch environmental stability level set, connects to all node data in the cloud road node platform, calls the real-time vibration amplitude, temperature fluctuation amplitude, and humidity fluctuation amplitude recorded in each node, compares whether they are greater than or equal to the minimum environmental standard value threshold corresponding to each batch, records the judgment result, retains the nodes that meet the three-dimensional conditions, and removes the nodes that do not meet the standard items to obtain a set of compliant nodes;

[0025] The structural map construction submodule extracts spatial location distance information pair by pair based on the number and geographical location attributes of each node in the compliant node set, determines whether there is continuity and spatial connection attributes, retains node pairs with connection relationships and establishes node index groups, removes node pairs without road connectivity information, aggregates and generates path connection structures according to node numbers, and obtains node structural map data groups.

[0026] The path segment extraction submodule extracts the environmental data corresponding to each node in the path segment based on the node structure map data group, extracts the minimum value of all nodes in the path segment among the three environmental standard values ​​of vibration, temperature and humidity, records the minimum guaranteed value set of the path segment, excludes path segments whose stability fluctuation range exceeds the preset range, and filters and retains segments according to node continuity and integrity to obtain a set of schedulable path segments.

[0027] As a further aspect of the present invention, the system further includes:

[0028] Based on the set of schedulable path segments, the cloud-based scheduling optimization module automatically matches all compliant path segments through the cloud platform, maps each batch of integrated agricultural product delivery to the path number, constructs a complete delivery scheduling path map, and generates optimized agricultural product delivery routes.

[0029] The optimized agricultural product delivery routes include the matching results of delivery batches and route numbers, the scheduling route diagram structure, and the final delivery route number.

[0030] As a further aspect of the present invention, the cloud scheduling optimization module includes:

[0031] The path segment matching submodule obtains the time window information and delivery target area location of each delivery batch based on the set of schedulable path segments, filters path segments with overlapping time periods according to the time correspondence, compares the spatial location identifiers of the delivery area center and the start and end nodes of the path segment, identifies path segments whose distance does not exceed the preset joint delivery range, and obtains the delivery path matching set.

[0032] The batch mapping submodule reads the basic attribute information of each path segment based on the list of candidate path segments for each delivery batch in the delivery path matching set, compares each path segment item by item, establishes a one-to-one mapping relationship between the path segment number and the delivery batch, and obtains a batch path mapping pair set.

[0033] The scheduling optimization submodule judges the continuity and structural smoothness of the path segments according to the batch path mapping set, rearranges the combination order of each path segment and delivery batch according to the smoothness, compares with the original allocation structure, updates the path allocation list, and obtains the optimization results of agricultural product delivery routes.

[0034] Compared with the prior art, the advantages and positive effects of the present invention are as follows:

[0035] This invention constructs a triple constraint rule of spatial direction, service range, and time window to accurately extract potential correlations between order nodes. Combined with the tolerance standards of agricultural product categories under multi-dimensional environmental indicators, it completes the differentiation of stability levels of delivery batches and filters environmental indicators of road nodes to ensure that the route selection process matches the actual transportation needs of agricultural products. It realizes linkage matching from route generation to batch scheduling, solves the problem of lack of transportation environment adaptation mechanism in traditional route planning, avoids the uncertainty of human experience in route setting and scheduling, improves the rationality of route construction, the accuracy of batch fusion, and the intelligence of node scheduling, and enhances the safety of the route selection process and the timeliness guarantee of the agricultural product circulation chain. Attached Figure Description

[0036] Figure 1 This is a system flowchart of the present invention;

[0037] Figure 2 This is a flowchart of the spatial node construction module of the present invention;

[0038] Figure 3 This is a flowchart of the batch fusion recognition module of the present invention;

[0039] Figure 4 This is a flowchart of the stability level configuration module of the present invention;

[0040] Figure 5 This is a flowchart of the node path filtering module of the present invention;

[0041] Figure 6 This is a flowchart of the cloud scheduling optimization module of the present invention. Detailed Implementation

[0042] To make the objectives, technical solutions, and advantages of this invention clearer, the invention will be further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative and not intended to limit the invention.

[0043] In the description of this invention, it should be understood that the terms "length," "width," "upper," "lower," "front," "rear," "left," "right," "vertical," "horizontal," "top," "bottom," "inner," and "outer," etc., indicating orientation or positional relationships, are based on the orientation or positional relationships shown in the accompanying drawings and are only for the convenience of describing the invention and simplifying the description, and do not indicate or imply that the device or element referred to must have a specific orientation, or be constructed and operated in a specific orientation, and therefore should not be construed as a limitation of the invention. Furthermore, in the description of this invention, "a plurality of" means two or more, unless otherwise explicitly specified.

[0044] Please see Figure 1 A cloud-based delivery route optimization system includes:

[0045] The spatial node construction module obtains the delivery latitude and longitude and time node of each agricultural product order, calculates whether the cosine value of the angle between the spatial direction vectors between nodes exceeds the path direction cosine similarity threshold (set ≥0.93, referring to the Dijkstra algorithm optimization standard), calculates whether the Euclidean distance between nodes is lower than the service radius threshold (set ≤5km according to the "Urban Logistics Distribution Service Specification"), and combines whether the delivery time difference is less than the time window grace period (set ±15min according to ISO15638-5:2020). It marks the node combination that simultaneously satisfies the three constraints and generates a set of nodes associated with direction and time distance.

[0046] The batch fusion identification module is based on the set of direction and time distance related nodes. It performs batch merging and marking on order nodes, establishes a mapping set between order number and batch identifier, counts the number of orders in each batch, performs unique number identification processing on each batch, and generates a fusion delivery batch structure set.

[0047] The stability level configuration module retrieves the environmental tolerance standards for each batch of agricultural products in terms of vibration, temperature, and humidity based on the integrated delivery batch structure set. It extracts the maximum requirements by taking the intersection of all standard values, constructs stability level differentiation standards according to the three dimensions, and assigns them to batches to generate a set of batch environmental stability levels.

[0048] The node path filtering module is based on the environmental tolerance standards corresponding to each batch in the batch environmental stability level set. It connects with all node data in the cloud road node platform, sequentially screening whether the vibration amplitude and temperature and humidity fluctuations meet the required minimum stability level. It excludes all nodes that do not meet the requirements, retains compliant nodes, constructs a path connection map, extracts path segments with complete structure, and generates a set of schedulable path segments.

[0049] The cloud-based scheduling optimization module is based on a set of schedulable path segments. It automatically matches all compliant path segments through the cloud platform, maps each batch of integrated agricultural product delivery to the path number, constructs a complete delivery scheduling path map, and generates optimized agricultural product delivery routes.

[0050] The set of direction and time distance related nodes includes direction vector similarity identifiers, service radius compliance identifiers, and time window overlap identifiers. The fused delivery batch structure set includes the mapping relationship between order number and batch identifier, batch order quantity statistics, and batch unique identifier code. The batch environmental stability level set includes vibration tolerance level, temperature adaptability level, and humidity control level. The schedulable path segment set includes environmental level matching path nodes, structurally complete path segment information, and path segment connection relationships. The agricultural product delivery route optimization results include the matching results between delivery batches and path numbers, the scheduling path graph structure, and the final delivery route number.

[0051] Please see Figure 2 The spatial node construction module includes:

[0052] The node information extraction submodule extracts the delivery latitude and longitude and delivery time node based on the delivery information in each agricultural product order, parses the latitude and longitude values, checks the data integrity and format consistency, and generates a spatiotemporal information group of delivery nodes.

[0053] Based on the delivery information in each agricultural product order, the delivery latitude and longitude and delivery time need to be extracted one by one. In the implementation process, the structured order information table is first read, and the order number, delivery longitude, delivery latitude, and delivery time fields are retrieved one by one. The data type of each field is validated to ensure that the latitude and longitude are in floating-point format and the delivery time is in a standardized time format, such as 10:00, 10:12, etc., and then converted into a unified timestamp format for subsequent calculations. After the field validation is completed, the longitude and latitude are extracted sequentially and combined into a two-dimensional coordinate vector to represent the delivery location. Referring to the following order data example, the coordinate points are (116.391, 39.907), (116.393, 39.908), and (116.398, 39.909). (116.402, 39.910) After obtaining the spatial coordinates, the data integrity is further checked. Data with missing fields is removed. For example, if a delivery time is not filled in for an order, the data will not participate in the subsequent node calculation. Then, the format consistency is verified to confirm that all nodes use the same unit and adopt a unified format. If there are abnormal string cases in the longitude, such as "E116.39", regular expression replacement is required to unify it into decimal form. A node information table is established in the verified data set to record the number, spatial location and timestamp of all valid delivery nodes, which serves as the basic data set for subsequent spatial matching and time distance comparison. In this example, 4 valid nodes are extracted, and the specific information is shown below. Finally, the spatiotemporal information group of delivery nodes is generated.

[0054] Table 1 Delivery Node Data Table

[0055] Order Number Delivery Longitude Delivery Dimensions Delivery time A001 116.391 39.907 10：00 A002 116.393 39.908 10：12 A003 116.398 39.909 10：25 A004 116.402 39.910 10：45

[0056] As shown in Table 1, all four nodes listed in the data meet the requirements of field integrity and format consistency, providing an initial basis for spatial node construction and time interval matching.

[0057] The spatial relationship determination submodule calculates the cosine of the angle between the spatial direction vectors based on the latitude and longitude coordinates of each pair of nodes in the spatiotemporal information group of the delivery node, filters node combinations with a cosine value not lower than the path direction cosine similarity threshold, calculates the Euclidean distance of the corresponding node pair, filters node combinations with an Euclidean distance not higher than the service radius threshold, and cross-compares to generate directional distance matching node pairs.

[0058] Based on the latitude and longitude coordinates of each pair of nodes in the delivery node spatiotemporal information group, the cosine of the angle between their spatial direction vectors needs to be calculated, and the reference path direction is set as a unit vector extending northeastward. Calculate the direction vector for each node pair. For example, the direction vector from A001 to A002 is (116.393-116.391, 39.908-39.907) = (0.002, 0.001). The formula for calculating the cosine of the corresponding angle is:

[0059]

[0060] Substituting the data, we get:

[0061]

[0062] This value is greater than the set path direction cosine similarity threshold of 0.93, therefore the node pair A001-A002 meets the direction similarity requirement. The cosine values ​​of nodes A001-A003, A001-A004, etc., are then evaluated sequentially, retaining only those greater than 0.93. Next, the Euclidean distance of each node pair is calculated using the formula:

[0063]

[0064] Taking A001-A002 as an example:

[0065] ;

[0066] Converting the units to kilometers, considering that each degree of latitude is approximately 111 kilometers, then:

[0067] d 12 =0.002236×111≈0.248km;

[0068] Since the value is below the service radius threshold of 5km, the node pair is retained. Other node combinations are compared in turn, and node pairs that simultaneously satisfy the directional similarity and Euclidean distance constraints are selected. These combinations are recorded as the basis dataset for subsequent time consistency judgments, and finally, directional distance matching node pairs are generated.

[0069] The direction and time distance matching submodule calculates the delivery time difference between nodes based on the delivery time of each pair of nodes in the direction and distance matching node pair, judges it against the time window grace period, filters the node combination that meets the time difference condition, and obtains the set of direction and time distance related nodes.

[0070] Based on the delivery time of each pair of nodes in the directional distance matching node pair, the time difference is calculated. For example, if the time of A001 is 10:00 and the time of A002 is 10:12, the time difference is 12 minutes. It is determined whether it is within the ±15 minute time window grace period. Since 12 minutes is less than 15 minutes, A001-A002 can be judged as a time-consistent combination. The delivery time of A001-A003 is 10:25, and the time difference is 25 minutes. It does not meet the time window requirement, so this combination is eliminated. The above operation is repeated to traverse and compare the remaining node pairs, retaining all node combinations that meet the time window grace requirement. Based on these combinations, their spatial direction and Euclidean distance attributes are integrated as triple constraints to establish a consistent node set with consistent delivery direction, close spatial distance, and concentrated time distribution. Finally, the set of nodes associated with direction and time distance is obtained.

[0071] Please see Figure 3 The batch fusion recognition module includes:

[0072] The node batch merging submodule extracts the order number information and associated node combination structure within the set based on the direction and time distance associated node set. It constructs a mutual belonging relationship graph for node combinations with cross pointing and establishes an order merging identifier set by grouping any node in the graph structure with all connected nodes in the same group.

[0073] Based on the set of nodes associated with direction and time, all matched order numbers and their combinations are extracted. Before merging, a batch graph structure needs to be constructed, where each node represents an order number and edges represent the direction and time compatibility between nodes. After establishing the graph structure by traversing all node combinations, each connected subgraph is identified. Each connected subgraph represents a batch set that can be merged. For example, A001 and A002 are direction and time matching objects, and A002 and A003 are connected. Therefore, A001, A002, and A003 are classified into the same batch B01. At the same time, a depth-first traversal is performed on the graph structure. During the traversal, all node numbers in the current path are recorded, and the set corresponding to the path is assigned a unified batch label. In the example, if nodes A004-A006 form another independent connected graph, they are merged into batch B02. During the execution, duplicate checks are performed on each set structure to ensure that each order belongs to only one batch. After marking, a one-to-one correspondence set between order numbers and batch labels is established, and finally, the order merging identifier set is obtained.

[0074] The batch order counting submodule counts the number of order nodes belonging to each batch based on the order merging identifier set, extracts the set of order numbers corresponding to each batch identifier, performs a length detection operation, records the number of orders, and constructs a batch order distribution array based on the quantity distribution to generate batch order quantity distribution data.

[0075] Based on the order merge identifier set, all order nodes are extracted and their quantities are counted by batch. During the counting process, the batch label is used as the index key to retrieve the number of nodes under the same label one by one, constructing a batch statistical mapping dictionary. For example, batch B01 contains 3 order nodes, B02 contains 5 nodes, and B03 contains only 1 node. A simple counter is used to accumulate and sum the matching results for each time. After the mapping is constructed, the number of corresponding nodes under all batches is extracted and formed into a length sequence, presented in the form of the sequential set {3,5,1}. The statistical results are used to measure the task size of each batch. This information will be used in the calculation and analysis when generating batch identifier values ​​in the future, and finally the batch order quantity distribution data is obtained.

[0076] The batch number generation submodule assigns a unique identifier to each batch based on the batch order quantity distribution data, using the following formula:

[0077]

[0078] The process involves calculating and obtaining the merged batch identifier value, binding the merged batch identifier value with the batch number, and establishing a merged delivery batch structure set, where B... k Indicates the batch identifier value, w ik L represents the spatial adjacency weight of the i-th order node in the k-th batch. ik d represents the path length between the i-th node and its predecessor node, in kilometers. ik Let represent the cosine value of the direction vector between the i-th node and the batch center. T represents the average value of the direction cosine of all nodes in the k-th batch. ik S represents the time difference between the delivery timestamp of the i-th node and the timestamp of the batch center node, in minutes. k n represents the standard deviation of all time differences in the k-th batch, in minutes. k This represents the number of order nodes in the k-th batch;

[0079] Based on the batch order quantity distribution data, a unique number is assigned to each batch and a fusion identifier value is generated. The spatial adjacency weight w of the order node is introduced into the generation of the fusion identifier. ik Path distance L ik , Directional consistency offset Time difference T ik and the standard deviation of the time difference S k The indicators are used as input parameters, and the calculation is performed using formulas.

[0080] Taking batch B01 as an example, assuming it includes order nodes A001, A002, and A003, substituting the parameter values ​​from Table 4, the expressions for each item are calculated as follows:

[0081] For A001:

[0082]

[0083] Term value = 1.0693 × 0.8013 = 0.8567;

[0084] For A002:

[0085]

[0086] Term value = 0.9162 × 1.2020 = 1.1015;

[0087] For A003:

[0088]

[0089] Term value = 1.12 × 1.6023 = 1.7946;

[0090] Summing the three terms, we get:

[0091] B 01 =|0.8567+1.1015+1.7946|=3.7528;

[0092] This value is the fusion batch identifier value of batch B01, representing the size of the aggregation characteristics of this batch in terms of comprehensive spatial adjacency, path distance, directional consistency and temporal concentration. Subsequently, this value will be bound to batch number B01 and used in the subsequent delivery system number index to finally establish the fusion delivery batch structure set.

[0093] Table 2 Calculation Parameters for Fusion Batch Identifier

[0094]

[0095] As shown in Table 2, the fusion value obtained after the actual values ​​of each node parameter and the substitution calculation is 3.7528. This result indicates that the matching quality of the current batch in terms of direction, distance and time interval is within a reasonable range, and its value can be used for batch sorting and priority calculation.

[0096] The merged batch identifier value is used to measure the aggregation strength of an order batch in four dimensions: spatial adjacency, path structure, directional consistency, and temporal concentration. It is a comprehensive expression of the delivery attributes of all order nodes within the batch. The larger the value, the more compact the geographical distribution, the more coherent the path connections, the more consistent the directional direction, and the more concentrated the delivery time. Therefore, this value can not only be used as a basis for batch numbering, but also to evaluate the rationality of batch fusion and execution priority. In actual scheduling, the merged batch identifier value has uniqueness and stability and can be bound to the batch code in the system to assist in delivery route arrangement, batch capacity allocation, and differentiation between multiple batches.

[0097] The formula first iterates through the participation of all nodes in the summation structure, taking each order node as a unit. The summation symbol ∑ is used to accumulate the contribution values ​​of all nodes within the batch, thus constructing the overall identifier strength of the batch. Within the summation term, the numerator is determined by the spatial adjacency weight w. ik Distance L from the path ik Multiplication is used to reflect the local influence of a node in the batch path. Spatial weight measures the tightness of connection, and path length reflects the actual distance traveled. Therefore, multiplying the two can quantify the structural value of a node at the spatial level. This product is normalized by dividing by a factor equal to the absolute value of the directional offset plus one. This represents the difference between the node's direction vector and the average of the batch directions. A smaller difference indicates greater consistency. The increment operation avoids division by zero errors and controls the penalty for the overall offset's contribution to the structure. Subsequently, this structure is compared with the square root of the time factor. The multiplication involves a time difference term to measure how close the node is to the batch core in terms of timeliness. The square root operation is used to mitigate the impact of large time differences, preventing them from causing exponential inflation of the overall batch value. This is further explained by comparing the multiplication with the batch standard deviation S. k The ratio is normalized to enhance the model's sensitivity to temporal concentration. Finally, the outer layer encapsulates all summation results using absolute values ​​to keep the batch identifier value positive, facilitating subsequent sorting and comparison. The entire formula combines multiple dimensions such as spatial structure, directional consistency, and temporal synchronization into a fusion metric based on physical meaning. In the computational structure, multiplication expresses interaction effects, division expresses penalty or constraint mechanisms, and square roots adjust nonlinear responses, reflecting the quantitative integration of the compactness and consistency characteristics of the batch's internal structure.

[0098] Please see Figure 4 The stability level configuration module includes:

[0099] The category environmental standard extraction submodule retrieves the corresponding agricultural product categories in each batch according to the integrated delivery batch structure set, obtains the environmental tolerance standard values ​​corresponding to the category under the three dimensions of vibration, temperature and humidity, matches and retrieves according to the category number and batch index, groups and stores them by batch number, and generates a batch category environmental standard dataset.

[0100] Based on the integrated delivery batch structure set, the corresponding agricultural product categories within each batch are retrieved one by one. Environmental tolerance parameters for each category are extracted under vibration, temperature, and humidity dimensions. The vibration dimension is expressed in meters per square second (m / s²), the temperature dimension in degrees Celsius, and the humidity dimension in percentage (relative humidity). Within each batch, a cyclical search is performed by category number. The three-dimensional environmental parameters corresponding to F01, F02, and F03 are retrieved from the vibration resistance standard library, suitable temperature range library, and humidity limit table, and these three data points are assigned to the corresponding fields in the table structure. If a category number has no matching entry during the search, it is marked as missing; if a duplicate number exists, the latest record is used to overwrite it. Finally, using the batch number as an index, the three parameters of different categories are merged into the same batch group structure, constructing a key-value correspondence between batch number and the three-dimensional parameter set. For example, batch B01 contains three agricultural products, F01 to F03, and their corresponding environmental parameters are as follows: F01 is 2.5 m / s². 2 At 5.0℃ and 80% RH, the FO2 velocity is 2.2 m / s. 2 At 7.0℃ and 85% RH, the F03 velocity is 2.4 m / s. 2 The environmental standards for each batch of products were recorded at 6.0℃ and 78%RH and are listed in Table 5.

[0101] Table 3 Environmental Standard Values ​​for Batch Categories

[0102]

[0103]

[0104] As shown in Table 3, the environmental standard values ​​for each category within batch B01 have been fully extracted and organized into the table.

[0105] The standard intersection acquisition submodule is based on the batch category environmental standard dataset, combining the vibration, temperature, and humidity standard values ​​corresponding to different categories in each batch, and using the following formula:

[0106]

[0107] The batch environment intersection fitness value is obtained by calculation; where E k This represents the fitness value of the environmental intersection in the k-th batch. G represents the minimum value in the set of maximum environmental standard values ​​for all categories under the j-th environmental dimension in the k-th batch. ajk This represents the environmental standard value of the a-th category in the k-th batch on the j-th environmental dimension, with units of m / s. 2 ℃, %RH N represents the average standard value of the j-th dimension across all categories within the batch. k Indicates the quantity of each category within this batch;

[0108] Based on the batch category environmental standard dataset, for the three categories F01 to F03 in batch B01, the maximum value was extracted and the minimum upper bound was obtained for the standard value set under three-dimensional parameters. Simultaneously, the deviation between each category parameter and the batch average was squared and the root mean square was calculated. This operation was performed on all three dimensions to obtain the difference fusion value for each dimension. In the first dimension, vibration, the maximum value set is {2.5, 2.2, 2.4}, where the maximum value is 2.5, the minimum upper bound is 2.2, and the mean is...

[0109]

[0110] The three deviations are as follows:

[0111] F01: |2.5-2.3667| 2 =0.0177;

[0112] F02: |2.2-2.3667| 2 =0.0278;

[0113] F03: |2.4-2.3667| 2 =0.0011;

[0114] Root mean square:

[0115]

[0116] The result of this dimension calculation is: 2.2 - 0.1245 = 2.0755;

[0117] In the second temperature dimension, the set of maximum values ​​is {5.0, 7.0, 6.0}, the minimum upper bound is 5.0, and the mean is:

[0118]

[0119] The corresponding sum of squared deviations is:

[0120] (5.0-6.0) 2 +(7.0-6.0) 2 +(6.0-6.0) 2 =1+1+0=2;

[0121]

[0122] The final dimension result is: 5.0 - 0.8165 = 4.1835;

[0123] In the third dimension of humidity, the set of maximum values ​​is {80, 85, 78}, the minimum upper bound is 78, and the average value is:

[0124]

[0125] The corresponding sum of squared deviations is:

[0126] (80-81) 2 +(85-81) 2 +(78-81) 2 =1 + 16 + 9 = 26;

[0127] The root mean square is:

[0128]

[0129] The dimensional result is: 78 - 2.9439 = 75.0561;

[0130] The sum of the final three dimensions is the batch intersection fitness:

[0131] E 01 =2.0755 + 4.1835 + 75.0561 = 81.3151;

[0132] This value represents the environmental co-fit value of batch B01 after fusion under three-dimensional parameters, and finally obtains the batch environment intersection fitness value.

[0133] The batch environmental intersection fitness value is a quantitative indicator used to measure the range of environmental conditions that all agricultural product categories in a delivery batch can jointly withstand under the three environmental dimensions of vibration, temperature, and humidity. The larger the value, the smaller the parameter differences among the categories within the batch in each environmental dimension, the more concentrated the overall adaptability, and the more suitable it is for storage and transportation under uniform conditions. This value comprehensively considers the maximum tolerable threshold of each category in each dimension, and by eliminating the degree of fluctuation in the distribution of internal standard values, it reflects the overall response characteristics of the batch to environmental stability in actual transportation control. Therefore, this value can be used as the basis for stability level classification to achieve the correspondence between batch and environmental control level.

[0134] The calculation logic of this formula aims to quantify the stability integration degree of each delivery batch under three environmental dimensions. First, within each dimension j, the smallest item in the set of maximum environmental standard values ​​for all categories a in the batch is extracted. This represents the most tolerant upper limit that all categories in that batch can jointly adapt to in that dimension. As a dimension to adapt to the upper bound benchmark; then the root mean square deviation term is introduced. To measure the degree of deviation of each category from the mean of the standard value in this dimension, the interference of positive and negative signs is eliminated by squaring, and then the root of the squared deviation is taken to make it comparable after standardization. This term is directly subtracted from the fitness upper limit, forming a deduction for the internal stability of the dimension. That is, if the distribution difference between categories is greater, the overall intersection value will be smaller. Finally, the above difference terms are accumulated in the three dimensions to reflect the parallelism of each dimension in the stability structure, while avoiding the dominant influence of a single dimension on the result. Therefore, the addition operation is used for comprehensive evaluation, the squaring and square root operations are used for standardization deviation measurement, and the subtraction operation reflects the stability penalty weight, finally forming the complete batch environment intersection fitness value E. k .

[0135] The hierarchy generation submodule combines the set environmental stability level classification benchmark, matches the batch environmental intersection fitness value with the level threshold, performs segment judgment on the intersection value of each batch according to the threshold range and assigns a level label to generate a batch environmental stability level set.

[0136] Based on the batch environment intersection fitness value of 81.3151, the system-defined environment level classification standard is invoked, with the following partitions: Level A is ≥85, Level B is 70 to 85, Level C is 50 to 70, and Level D is <50. The current intersection value is within the range of 70 to 85, and according to the segment judgment rule, it is classified as Level B. Level B is written as the label into the level identifier field of batch B01. At the same time, a mapping set structure between level number and batch number is constructed to finally obtain the batch environment stability level set.

[0137] Please see Figure 5 The node path filtering module includes:

[0138] The node environment filtering submodule, based on the minimum vibration tolerance standard value, minimum temperature tolerance standard value, and minimum humidity tolerance standard value corresponding to each batch in the batch environmental stability level set, connects to all node data in the cloud road node platform, calls the real-time vibration amplitude, temperature fluctuation amplitude, and humidity fluctuation amplitude recorded in each node, compares whether they are greater than or equal to the minimum environmental standard threshold value corresponding to each batch, records the judgment result, retains the nodes that meet the three-dimensional conditions, and removes the nodes that do not meet the standard items to obtain the set of compliant nodes;

[0139] Based on the minimum tolerance standards for vibration, temperature, and humidity for each batch in the batch environmental stability level set, and by connecting with the real-time environmental parameter data of all nodes in the cloud-based road node platform, the current vibration amplitude, temperature fluctuation, and humidity fluctuation parameters are extracted for nodes N01 to N05 respectively, and each parameter is compared with the minimum threshold of the batch to which that node belongs. For example, the vibration amplitude of node N01 is 1.8 m / s. 2 This is below the minimum vibration threshold of 2.0 m / s for its batch. 2 The nodes were deemed non-compliant. Node N02's temperature fluctuation was 2.8℃, lower than the batch's minimum temperature threshold of 3.0℃, and was therefore also deemed non-compliant. In the humidity fluctuation assessment, node N03's RH was 11%, lower than the batch's humidity threshold of 12%, and also did not meet the requirements. Node N04, although meeting the vibration and temperature thresholds, was excluded because its humidity was only 9% RH. Node N05's vibration was 1.9 m / s. 2 The threshold was not met. Of the five nodes, only node N03 met the vibration and temperature requirements, but the humidity requirement was not met. Therefore, it did not meet the screening criteria. The vibration threshold was set to 2.0 m / s. 2 The standard is based on the vibration control standards for equipment during agricultural product transportation. The temperature threshold of 3.0℃ is taken from the minimum fluctuation tolerance of short-distance transportation for each category. The humidity threshold of 12%RH is derived from the experimental tolerance of humidity-sensitive agricultural products to air disturbance fluctuations. After setting the comparison standard, the parameter comparison results before and after screening are recorded as follows:

[0140] Table 4. Node Environmental Parameter Judgment Table

[0141]

[0142] As shown in Table 4, if any node falls below the set threshold in any dimension, the final set of compliant nodes will be empty.

[0143] The structural map construction submodule extracts spatial distance information pair by pair based on the number and geographical location attributes of each node in the compliant node set, determines whether there is continuity and spatial connection attributes, retains node pairs with connection relationships and establishes node index groups, removes node pairs without road connectivity information, aggregates and generates path connection structures according to node numbers, and obtains node structural map data groups.

[0144] Based on the node set already selected from the node environment filtering results, the spatial coordinate identifier field of each node is read, the horizontal and vertical position values ​​are extracted, and the Euclidean distance between any two nodes is calculated. For example, if the coordinates of nodes N03 and N04 are (114.22, 30.31) and (114.25, 30.34) respectively, then their horizontal and vertical differences are 0.03 and 0.03 respectively. The distance is calculated by squaring these values, adding them together, and then taking the square root. If a node is less than the maximum connection radius of 0.1km, it is recorded as a connectable node pair. The remaining nodes are then screened sequentially. If the distance between a node and all other nodes exceeds the connection radius, it is considered an isolated node and discarded. At the same time, the node attribute fields are called to determine whether it has connectivity attributes. If any node pair is marked as "disconnected", the connection is not recorded. After the screening, a set structure of connection relationships is constructed, and the nodes are regrouped and identified as structural graph units according to the order of path segments. Each graph unit records the path segment length, number of nodes, and connectivity identifier by aggregating path segments. Paths with interruptions in the structure are grouped and cut into sub-segment structures, and the path segments that finally meet the continuous connection relationship are numbered and stored in the database to obtain the node structural graph data group.

[0145] The path segment extraction submodule extracts the environmental data corresponding to each node in the path segment based on the node structure map data group, extracts the minimum value of all nodes in the path segment among the three environmental standard values ​​of vibration, temperature and humidity, records the minimum guaranteed value set of the path segment, excludes path segments whose stability fluctuation range exceeds the preset range, and filters and retains segments according to node continuity and integrity to obtain a set of schedulable path segments.

[0146] Based on the path segment information recorded in the node structure map data set, three data fields were extracted for each path segment: vibration amplitude, temperature fluctuation, and humidity fluctuation. The minimum value of each field was then determined. Path segment P01 contains nodes N03 and N04, with vibration values ​​of 2.7 m / s. 2 With 2.3m / s 2 The minimum value is taken as 2.3 m / s. 2 The minimum guaranteed value set is recorded for the temperature range of 4.2℃ and 3.1℃, with 3.1℃ being the lowest. The minimum guaranteed value set is recorded for the humidity range of 11%RH and 9%RH, with 9%RH being the lowest. The number of nodes is recorded as 2. The total span of the path segment is calculated from the geographical location between the two nodes, and its length is approximately 0.042km. The node density is calculated by dividing the number of nodes by the length, i.e., 2÷0.042≈47.6 nodes / km. The start and end time identifier fields of the path segment are then read and marked as the scheduling period T1. Finally, the minimum value extracted from the path segment is compared with the maximum allowable fluctuation tolerance set by the batch environmental stability level. If the lower limit of the batch allowable humidity fluctuation range is 10%RH, then the minimum value of this path segment is 9%RH, which does not meet the standard, and this segment number is removed. If the number of path segment numbers retained is less than 1, then the map path group cannot form a usable segment, and the final set of schedulable path segments is empty.

[0147] Please see Figure 6 The cloud-based scheduling optimization module includes:

[0148] The path segment matching submodule is based on the set of schedulable path segments. It obtains the time window information and the location of the delivery target area for each delivery batch, screens path segments with overlapping time periods according to the time correspondence, compares the spatial location markers of the delivery area center and the start and end nodes of the path segment, identifies path segments whose distance does not exceed the preset joint delivery range, and obtains the delivery path matching set.

[0149] Based on the route segment numbers and corresponding time periods and geographical coordinates recorded in the set of schedulable route segments, the time window field and delivery area identifier of each delivery batch are read sequentially. When processing the first delivery batch, its planned departure time is read as 08:00 to 10:00 on June 30, 2024. This time period is compared with the time period field 06:00 to 09:00 of route segment P01 to determine time overlap. A 2-hour overlap interval is identified, classifying it as a time-matching route segment. Simultaneously, the coordinates of the delivery area center are read as 114.30°E, 30.50°N, and the coordinates of the start and end nodes of route segment P01 are 114.28°E, 30.48°N and 114.31°E, 30.52°N, respectively. The Euclidean distances between these two points and the area center are calculated as 0.028km and 0.036km, respectively, both less than [the required distance]. The path matching threshold is set to 0.05km. Path segment P01 is confirmed as a candidate path segment for this delivery batch, and its number is added to the path segment matching list for this batch. Subsequently, delivery batches 2, 3, etc., are processed, and the overlap and fit of the time and geographical distance fields are judged respectively. Path segment numbers with matching characteristics are filtered out and marked and included in the path candidate set for this batch. After processing each batch of delivery tasks, a matching structure between batch number and path segment number is established. This structure can be used for pairing determination before scheduling. In the overall operation, the path matching radius threshold is set to 0.05km, which is jointly formulated based on the shortest boundary distance of the path segment and the vehicle scheduling response time limit. The time window matching must meet the condition that the start and end overlap time is not less than 30 minutes. The matching process of path segment P01 and batch 1 is shown in Table 5. Finally, the delivery path matching set is obtained.

[0150] Table 5. Sample Table of Batch and Path Segment Matching

[0151]

[0152] As shown in Table 5, a path segment is determined to be a matching path segment only if it meets the set threshold requirements in both the time window and spatial distance fields.

[0153] The batch mapping submodule reads the basic attribute information of each path segment based on the list of candidate path segments for each delivery batch in the delivery path matching set, compares each path segment item by item, establishes a one-to-one mapping relationship between the path segment number and the delivery batch, and obtains the batch path mapping pair set.

[0154] Based on the candidate list of route segments for each delivery batch in the delivery route matching group, the attribute fields of each route segment are read one by one. First, the number of nodes for route segment P01 is extracted (6), for route segment P02 (9), and for P03 (7). The route span fields are recorded as 1.2km, 1.9km, and 1.5km respectively. The integrity field of node connection is identified through data identifiers. For example, all nodes in P01 are "connected," so the integrity score is 1. P02 has one node identified as "disconnected," so the integrity score is 0. The path number and batch number correspondence of each segment from P01 to P03 are called and compared. During the matching process, the node integrity field is judged first, and only route segments with an integrity score of 1 are retained. If any other path segments are found to be missing, the remaining path segments are removed. If multiple path segments are complete, the degree of fit between the number of nodes and the span field is further determined. For example, if the expected path span limit for batch B01 is 1.5km, and P02 exceeds the limit and is excluded as a candidate, then P01 and P03 are still candidates. Comparing the node density field, P01 has a density of 6 nodes / 1.2km ≈ 5.0 nodes / km, and P03 has a density of 4.6 nodes / km. Higher density indicates stronger data continuity. Finally, P01 is selected as the target path segment for batch B01, and its number is mapped to B01 to form a mapping pair, which is written into the path mapping table. The target path segment for batch B02 is P03, and after the mapping is established, it is written into the table. This completes the one-to-one correspondence binding operation between batches and path segments, and finally obtains the batch path mapping pair set.

[0155] The scheduling optimization submodule judges the continuity and structural smoothness of the path segments one by one based on the batch path mapping set, rearranges the combination order of each path segment and delivery batch according to the smoothness, compares with the original allocation structure, updates the path allocation list, and obtains the optimization results of agricultural product delivery routes.

[0156] Based on the pairing information in the batch path mapping set, the structural field information of the nodes contained in each path segment is read one by one. For the six nodes in path segment P01, the traffic status field and traffic condition identifier field are extracted item by item, which are "smooth", "slow", "smooth", "smooth", "slow", and "smooth" in sequence. The connection strength field between each node is recorded as strong, medium, strong, strong, medium, and strong. The structural density field is calculated from the average distance between each pair of adjacent nodes. For example, if the average distance between nodes is 200 meters, then the density field of the entire segment is 5 nodes / km. During the judgment process, it is identified that the segment has high continuity at the structural connection level. The traffic status is 4 smooth and 2 slow, with a smooth traffic rate of approximately 66.7%. When comparing the traffic status fields of other path segments, path segment P03 has 3 nodes that are "congested", and its smooth traffic rate is less than 50%. The structure density field is 3.9 nodes / km, which is lower than that of P01. Therefore, P01 is determined to be superior to other path segments in terms of structural continuity and traffic conditions. The order of the path segments corresponding to B01 is readjusted and updated to prioritize matching P01. B02 is not adjusted because only P03 is available. Finally, the sorting results are written into the scheduling result output table to form a new path allocation order and obtain the optimization results of agricultural product delivery routes.

[0157] The above are merely preferred embodiments of the present invention and are not intended to limit the present invention in any other way. Any person skilled in the art may make changes or modifications to the above-disclosed technical content to create equivalent embodiments that can be applied to other fields. However, any simple modifications, equivalent changes, and modifications made to the above embodiments based on the technical essence of the present invention without departing from the scope of the present invention shall still fall within the protection scope of the present invention.

Claims

1. A cloud computing-based delivery route optimization system, characterized in that, The system includes: The spatial node construction module obtains the latitude and longitude of delivery and time node in each agricultural product order, calculates whether the cosine value of the angle between the spatial direction vectors between nodes exceeds the path direction cosine similarity threshold, calculates whether the Euclidean distance between nodes is lower than the service radius threshold, combines whether the delivery time difference is less than the time window grace period, marks the node combination that simultaneously satisfies the three constraints, and generates a set of nodes associated with direction and time distance. Based on the set of direction and time distance related nodes, the batch fusion identification module performs batch merging and marking on order nodes, assigns a unique number to each batch, and generates a fusion delivery batch structure set. The stability level configuration module retrieves the environmental tolerance standards of each batch of agricultural products in the dimensions of vibration, temperature and humidity based on the integrated delivery batch structure set, constructs the stability level differentiation standard, and generates a set of batch environmental stability levels. Based on the batch environmental stability level, the node path filtering module connects to the node data in the cloud road node platform to screen whether the vibration amplitude and temperature and humidity fluctuations meet the minimum stability level, retains compliant nodes and extracts path segments with complete structure, and generates a set of schedulable path segments.

2. The cloud computing-based delivery route optimization system according to claim 1, characterized in that, The set of direction and time distance associated nodes includes direction vector similarity identifiers, service radius compliance identifiers, and time window overlap identifiers. The integrated delivery batch structure set includes the mapping relationship between order number and batch identifier, batch order quantity statistics, and batch unique identifier code. The set of batch environmental stability levels includes vibration tolerance level, temperature adaptability level, and humidity control level. The set of schedulable path segments includes environmental level matching path nodes, structurally complete path segment information, and path segment connection relationship.

3. The cloud-based delivery route optimization system according to claim 1, characterized in that, The spatial node construction module includes: The node information extraction submodule extracts the delivery latitude and longitude and delivery time node based on the delivery information in each agricultural product order, parses the latitude and longitude values, checks the data integrity and format consistency, and generates a spatiotemporal information group of delivery nodes. The spatial relationship determination submodule calculates the cosine of the angle between the spatial direction vectors based on the latitude and longitude coordinates of each pair of nodes in the delivery node spatiotemporal information group, filters node combinations with cosine values ​​not lower than the path direction cosine similarity threshold, calculates the Euclidean distance of the corresponding node pair, filters node combinations with Euclidean distance not higher than the service radius threshold, and cross-compares to generate directional distance matching node pairs. The direction and time distance matching submodule calculates the delivery time difference between nodes based on the delivery time of each pair of nodes in the direction and distance matching node pair, judges it against the time window grace period, filters the node combination that meets the time difference condition, and obtains the direction and time distance associated node set.

4. The cloud computing-based delivery route optimization system according to claim 1, characterized in that, The batch fusion identification module includes: The node batch merging submodule extracts the order number information and associated node combination structure within the set of nodes associated with direction and time distance based on the aforementioned set. It constructs a mutual belonging relationship graph for node combinations with cross-pointing relationships and establishes an order merging identifier set by grouping any node in the graph structure with all connected nodes in the same group. The batch order counting submodule counts the number of order nodes belonging to each batch based on the order merging identifier set, extracts the corresponding order number set under each batch identifier, performs a length detection operation, records the number of orders, and constructs a batch order distribution array based on the quantity distribution to generate batch order quantity distribution data. The batch number generation submodule assigns a unique number to each batch based on the batch order quantity distribution data, calculates and obtains the merged batch identifier value, binds the merged batch identifier value with the batch number, and establishes a merged delivery batch structure set.

5. The cloud-based delivery route optimization system according to claim 1, characterized in that, The stability level configuration module includes: The category environmental standard extraction submodule retrieves the corresponding agricultural product categories in each batch according to the integrated delivery batch structure set, obtains the environmental tolerance standard values ​​corresponding to the category under the three dimensions of vibration, temperature and humidity, performs matching and retrieval according to the category number and batch index, groups and stores them by batch number, and generates a batch category environmental standard dataset. The standard intersection acquisition submodule calculates and obtains the batch environment intersection fitness value based on the batch category environmental standard dataset and the three standard values ​​of vibration, temperature and humidity corresponding to different categories in each batch. The hierarchical structure generation submodule combines the set environmental stability level division benchmark, matches and judges the batch environmental intersection fitness value with the level threshold, judges the segment of each batch intersection value according to the threshold range and assigns a level label, and generates a batch environmental stability level set.

6. The cloud-based delivery route optimization system according to claim 1, characterized in that, The node path filtering module includes: The node environment filtering submodule, based on the minimum vibration tolerance standard value, minimum temperature tolerance standard value, and minimum humidity tolerance standard value corresponding to each batch in the batch environmental stability level set, connects to all node data in the cloud road node platform, calls the real-time vibration amplitude, temperature fluctuation amplitude, and humidity fluctuation amplitude recorded in each node, compares whether they are greater than or equal to the minimum environmental standard value threshold corresponding to each batch, records the judgment result, retains the nodes that meet the three-dimensional conditions, and removes the nodes that do not meet the standard items to obtain a set of compliant nodes; The structural map construction submodule extracts spatial location distance information pair by pair based on the number and geographical location attributes of each node in the compliant node set, determines whether there is continuity and spatial connection attributes, retains node pairs with connection relationships and establishes node index groups, removes node pairs without road connectivity information, aggregates and generates path connection structures according to node numbers, and obtains node structural map data groups. The path segment extraction submodule extracts the environmental data corresponding to each node in the path segment based on the node structure map data group, extracts the minimum value of all nodes in the path segment among the three environmental standard values ​​of vibration, temperature and humidity, records the minimum guaranteed value set of the path segment, excludes path segments whose stability fluctuation range exceeds the preset range, and filters and retains segments according to node continuity and integrity to obtain a set of schedulable path segments.

7. The cloud-based delivery route optimization system according to claim 1, characterized in that, The system also includes: Based on the set of schedulable path segments, the cloud-based scheduling optimization module automatically matches all compliant path segments through the cloud platform, maps each batch of integrated agricultural product delivery to the path number, constructs a complete delivery scheduling path map, and generates optimized agricultural product delivery routes. The optimized agricultural product delivery routes include the matching results of delivery batches and route numbers, the scheduling route diagram structure, and the final delivery route number.

8. The cloud-based delivery route optimization system according to claim 7, characterized in that, The cloud-based scheduling optimization module includes: The path segment matching submodule obtains the time window information and delivery target area location of each delivery batch based on the set of schedulable path segments, filters path segments with overlapping time periods according to the time correspondence, compares the spatial location identifiers of the delivery area center and the start and end nodes of the path segment, identifies path segments whose distance does not exceed the preset joint delivery range, and obtains the delivery path matching set. The batch mapping submodule reads the basic attribute information of each path segment based on the list of candidate path segments for each delivery batch in the delivery path matching set, compares each path segment item by item, establishes a one-to-one mapping relationship between the path segment number and the delivery batch, and obtains a batch path mapping pair set. The scheduling optimization submodule judges the continuity and structural smoothness of the path segments according to the batch path mapping set, rearranges the combination order of each path segment and delivery batch according to the smoothness, compares with the original allocation structure, updates the path allocation list, and obtains the optimization results of agricultural product delivery routes.

Images