Intelligent order allocation optimization method for freight platform based on large model

Abstract

The present invention discloses a method for optimizing intelligent order allocation of a freight platform based on a large model, and relates to the field of intelligent logistics technology. The method comprises: connecting to a freight platform to read a first freight order, establishing a first freight allocation resource list based on freight resource allocation constraints, combining the first freight order and the list to output an allocation result through a freight order allocation model, calculating fitness according to an order allocation fitness function, judging whether the constraints are met, and generating optimization instructions if they are not met, optimizing the first freight order allocation result according to the instructions and related content, and generating an optimization result. The method solves the technical problem that the accuracy of matching freight orders and freight resources in existing freight platforms is poor, resulting in low fitness of freight order allocation, and achieves the technical effect of improving the operational efficiency of the freight platform by realizing accurate matching of freight resources, scientifically and rationally allocating orders, and ensuring high fitness.

Description

Technical Field

[0001] The present application relates to the field of intelligent logistics technology, and in particular to an intelligent order allocation optimization method for a freight platform based on a large model. Background Art

[0002] In the order allocation scenario of freight platforms, the rationality of order allocation during the peak period of logistics business is particularly prominent. Efficient allocation of freight resources and better meeting the transportation needs of different orders have become a crucial link in improving the operational efficiency of freight platforms. Traditional freight platform order allocation is often relatively simple and localized, relying only on manual experience or simple rules. It lacks overall control over freight resources and lacks in-depth analysis and utilization of historical order data. It is difficult to fully and accurately evaluate the matching degree between different resources and orders. There are unreasonable situations in resource allocation, resulting in insufficient resources for some orders and idle resources for some. The formulation of order allocation plans is relatively fixed and cannot cope well with the ever-changing actual situation of freight business.

[0003] In the current relevant technologies, the matching accuracy between freight orders and freight resources in freight platforms is poor, which leads to the technical problem of low adaptability of freight order allocation. Summary of the invention

[0004] The present application provides a method for optimizing intelligent order allocation for a freight platform based on a large model, connects to a freight platform to read a first freight order, establishes a first freight allocation resource list based on freight resource allocation constraints, outputs an allocation result through a freight order allocation model in combination with the first freight order and the list, calculates fitness according to an order allocation fitness function, determines whether the constraints are met, generates optimization instructions if not, optimizes the first freight order allocation result according to the instructions and related content, and generates an optimization result, thereby achieving a technical effect of improving the operational efficiency of the freight platform by achieving accurate matching of freight resources, scientific and reasonable allocation of orders, and ensuring high fitness.

[0005] This application provides a method for optimizing intelligent order allocation for a freight platform based on a large model, including:

[0006] A freight order management module connected to a freight platform reads a first freight order; based on a freight resource registration constraint, performs resource registration on a freight resource pool of the freight platform according to the first freight order, and establishes a first freight registration resource list; based on the first freight order and the first freight registration resource list, according to a freight order allocation model, outputs a first freight order allocation result; based on the first freight order allocation result, calculates the first order allocation fitness according to an order allocation fitness function; determines whether the first order allocation fitness satisfies the order allocation fitness constraint; if the first order allocation fitness does not satisfy the order allocation fitness constraint, generates a first order allocation optimization instruction; based on the first order allocation optimization instruction, optimizes and allocates the first freight order allocation result according to the order allocation fitness function, the order allocation fitness constraint, the first freight registration resource list and the first freight order, and generates a first freight order allocation optimization result.

[0007] The intelligent order allocation optimization method for the freight platform based on a large model proposed in this application is to first connect to the freight platform to read the first freight order, establish a first freight matching resource list based on the freight resource matching constraints, and output the allocation result through the freight order allocation model in combination with the first freight order and the list. The fitness is calculated according to the order allocation fitness function to determine whether the constraints are met. If not, an optimization instruction is generated. The first freight order allocation result is optimized according to the instruction and related content to generate an optimization result. By realizing accurate matching of freight resources, scientific and reasonable allocation of orders, and ensuring high fitness, the technical effect of improving the operational efficiency of the freight platform is achieved. BRIEF DESCRIPTION OF THE DRAWINGS

[0008] In order to more clearly illustrate the technical solution of the embodiment of the present invention, the accompanying drawings of the embodiment of the present invention will be briefly introduced below. A flow chart is used in the present application to illustrate the operations performed by the system according to the embodiment of the present application. It should be understood that the preceding or following operations are not necessarily performed accurately in order. On the contrary, various steps can be processed in reverse order or simultaneously as needed. At the same time, other operations can also be added to these processes, or one or more operations can be removed from these processes.

[0009] Figure 1 A flow chart of a method for optimizing intelligent order allocation for a freight platform based on a large model provided in an embodiment of the present application.

[0010] Figure 2 A schematic diagram of the first freight order allocation optimization result generation process of the large model-based freight platform intelligent order allocation optimization method provided in an embodiment of the present application. DETAILED DESCRIPTION

[0011] The above description is only an overview of the technical solution of the present application. In order to more clearly understand the technical means of the present application, it can be implemented in accordance with the contents of the specification. In order to make the above and other purposes, features and advantages of the present application more obvious and easy to understand, the specific implementation methods of the present application are listed below.

[0012] In order to make the purpose, technical solutions and advantages of the present application clearer, the present application will be further described in detail below in conjunction with the accompanying drawings. The described embodiments should not be regarded as limiting the present application. All other embodiments obtained by ordinary technicians in the field without making creative work are within the scope of protection of this application.

[0013] In the following description, reference is made to "some embodiments", which describe a subset of all possible embodiments, but it is understood that "some embodiments" may be the same subset or different subsets of all possible embodiments, and may be combined with each other without conflict, and the terms "first\second" involved are merely to distinguish similar objects and do not represent a specific ordering of objects. The terms "including" and "having" and any variations are intended to cover non-exclusive inclusions, for example, a process, method, system, product, or server that includes a series of steps or units is not necessarily limited to those steps or units clearly listed, but may include other steps or modules that are not clearly listed or inherent to these processes, methods, products, or devices. Unless otherwise defined, all technical and scientific terms used herein have the same meaning as those generally understood by technicians in the technical field of this application. The terms used herein are for the purpose of describing the embodiments of the present application only.

[0014] The present application embodiment provides a method for optimizing intelligent order allocation for a freight platform based on a large model, such as Figure 1 As shown, the method includes:

[0015] Step S100, connect to the freight order management module of the freight platform, and read the first freight order. Specifically, establish a connection with the freight order management module of the freight platform through an interface or communication protocol, clearly read the first freight order as the target, determine the order identifier or index, read detailed information including the type of goods, quantity, size, weight, origin and destination address information, transportation time requirements, special transportation requirements, and customer contact information, and then verify the read data to ensure completeness and accuracy. If there are any problems, take corresponding measures, and organize the data at the same time, such as standardizing the address format, converting the transportation time requirements into specific time intervals, etc., to provide basic data for subsequent resource registration and order allocation.

[0016] Step S200, based on the freight resource registration constraints, the freight resource pool of the freight platform is registered with resources according to the first freight order, and a first freight registration resource list is established. Specifically, first determine the freight resource registration constraints, including freight driver qualification requirements, freight vehicle performance requirements, transportation distance restrictions and transportation time requirements, then analyze the first freight order requirements, determine the type and characteristics of the goods, quantity and volume, starting point and destination location to calculate the transportation distance and possible routes and evaluate the transportation time, while considering the special requirements of customers, and then register the resources of the freight resource pool of the freight platform. For the freight driver resource pool, select according to the qualification requirements and consider time availability; for the freight vehicle resource pool, select vehicles with appropriate load capacity and space and corresponding transportation functions according to the quantity and volume of the goods, and check the maintenance status and performance of the vehicles. Finally, organize the freight drivers and freight vehicles that meet the requirements of the first freight order into the first freight registration resource list, record detailed information, and provide a range of optional resources for subsequent order allocation.

[0017] In one possible implementation, based on the freight resource registration constraint, the freight resource pool of the freight platform is registered with resources according to the first freight order, and a first freight registration resource list is established. Step S200 further includes step S210, and the freight resource registration constraint includes a freight driver registration threshold and a freight vehicle registration threshold. Specifically, the freight resource registration constraint includes a freight driver registration threshold and a freight vehicle registration threshold, and clarifies the standard for resource registration. The setting of the registration threshold is based on the business needs, experience, and requirements for freight quality and efficiency of the freight platform. For example, the freight driver registration threshold may include comprehensive considerations of driving experience requirements, safe driving record requirements, service evaluation level requirements, etc. The freight vehicle registration threshold may involve the vehicle's load weight range, vehicle age limit, maintenance status standards, etc. The threshold provides a specific measurement standard for the subsequent registration of the freight driver resource pool and the freight vehicle resource pool.

[0018] In step S220, the freight resource pool includes a freight driver resource pool and a freight vehicle resource pool. Specifically, the freight resource pool includes a freight driver resource pool and a freight vehicle resource pool. The freight driver resource pool stores relevant information of many freight drivers, such as personal basic information, driving qualifications, work experience, service evaluation, etc. The freight vehicle resource pool contains information of various types of freight vehicles, including vehicle model, load capacity, special functions (such as refrigeration function), vehicle age, maintenance records, etc. Understanding the composition of the freight resource pool is helpful for targeted resource matching.

[0019] Step S230, based on the freight driver registration threshold, the freight driver resource pool is registered according to the first freight order to generate a first freight registered driver resource pool. Specifically, based on the freight driver registration threshold, the freight driver resource pool is registered according to the first freight order to generate a first freight registered driver resource pool. First, the characteristics and requirements of the first freight order are analyzed, such as transportation time requirements, special cargo handling requirements, etc. Then, qualified drivers are screened from the freight driver resource pool against the freight driver registration threshold. For example, if the order requires expedited transportation, then those drivers with rich expedited transportation experience and good evaluation are selected. The driver's driving experience, recent safe driving record, etc. are also strictly screened to ensure that only drivers who meet the registration threshold can enter the first freight registered driver resource pool.

[0020] Step S240, based on the freight vehicle registration threshold, the freight vehicle resource pool is registered according to the first freight order to generate a first freight registered vehicle resource pool. Specifically, based on the freight vehicle registration threshold, the freight vehicle resource pool is registered according to the first freight order to generate a first freight registered vehicle resource pool, first analyze the properties of the goods in the first freight order, such as the quantity and volume of the goods, whether special transportation conditions are required, etc., and select a suitable vehicle from the freight vehicle resource pool according to the freight vehicle registration threshold. If the goods are large and the volume is large, select a vehicle with a large load capacity and spacious space. For goods that need to be refrigerated, only vehicles with refrigeration functions and good maintenance can be selected to enter the first freight registered vehicle resource pool.

[0021] Step S250, sorting the first freight registration driver resource pool and the first freight registration vehicle resource pool to generate the first freight registration resource list. Specifically, sorting the first freight registration driver resource pool and the first freight registration vehicle resource pool to generate the first freight registration resource list, integrating the screened freight driver and freight vehicle information to form a comprehensive resource list, the list provides a specific range of optional resources for subsequent freight order allocation, and facilitates more accurate allocation according to the actual needs of the order. In the list, the qualifications and characteristics of each freight driver, as well as the performance and status of each freight vehicle can be recorded in detail, so as to make the best choice in the allocation process.

[0022] In a possible implementation, based on the freight driver registration threshold, the freight driver resource pool is registered according to the first freight order to generate a first freight registered driver resource pool, and step S230 further includes step S231, performing freight driver demand analysis based on the first freight order to generate first order driver qualification requirements and first order driver time requirements. Specifically, the freight driver demand analysis is performed based on the first freight order to generate first order driver qualification requirements and first order driver time requirements, analyze the specific requirements of the first freight order, such as the type, quantity, and special transportation requirements of the goods, determine the qualifications that the required freight driver should have, such as whether a specific type of driver's license (such as a dangerous goods transportation license) is required, whether rich long-distance transportation experience is required, etc., so as to generate the first order driver qualification requirements, consider the transportation time requirements of the order, delivery deadline and other factors, determine the time requirements for freight drivers, such as requiring drivers to be able to perform transportation tasks within a specific time period, and generate the first order driver time requirements.

[0023] Step S232, connect to the freight driver resource pool, read the first driver qualification information and the first driver freight schedule of the first freight driver. Specifically, connect to the freight driver resource pool, read the first driver qualification information and the first driver freight schedule of the first freight driver, and establish a connection with the freight driver resource pool to obtain the detailed information of each freight driver stored therein. For the first freight driver, read his qualification information, including driving age, driver's license type, past transportation record, service evaluation, etc., read the first driver freight schedule, and understand the driver's scheduled freight time, leave time, rest time, etc., so as to evaluate his time availability.

[0024] Step S233, based on the first order driver qualification requirements and the first driver qualification information, an adaptation evaluation is performed to obtain the first driver qualification adaptation coefficient. Specifically, based on the first order driver qualification requirements and the first driver qualification information, an adaptation evaluation is performed to obtain the first driver qualification adaptation coefficient, and the driver qualification requirements required by the first freight order are compared with the actual qualification information of the first driver one by one. For example, if the order requires the driver to have more than five years of long-distance transportation experience, and the first driver has exactly six years of long-distance transportation experience, then the degree of adaptation in this dimension is relatively high. By setting a certain evaluation algorithm, each qualification dimension is quantitatively evaluated, and finally the first driver qualification adaptation coefficient is obtained, which reflects the degree of matching between the first driver and the first freight order in terms of qualifications.

[0025] Step S234, based on the time demand of the driver of the first order and the freight schedule of the first driver, an adaptation evaluation is performed to obtain the freight time adaptation coefficient of the first driver. Specifically, based on the time demand of the driver of the first order and the freight schedule of the first driver, an adaptation evaluation is performed to obtain the freight time adaptation coefficient of the first driver, and the time demand of the driver of the first freight order is compared with the freight schedule of the first driver. If the order requires the driver to have time to perform the transportation task within the next three days, and the first driver's freight schedule is checked to find that there are no scheduled freight tasks within these three days and it is not during leave or rest time, then the adaptation degree in the time dimension is high, and the time adaptation degree is quantified to obtain the freight time adaptation coefficient of the first driver, which reflects the degree to which the first driver meets the requirements of the first freight order in time.

[0026] Step S235, weighted calculation is performed on the first driver qualification adaptation coefficient and the first driver freight time adaptation coefficient based on the freight driver matching weight condition to obtain the first freight driver matching coefficient. Specifically, weighted calculation is performed on the first driver qualification adaptation coefficient and the first driver freight time adaptation coefficient based on the freight driver matching weight condition to obtain the first freight driver matching coefficient. The weights of the qualification adaptation coefficient and the freight time adaptation coefficient are determined according to the business focus and strategy of the freight platform. For example, if the platform pays more attention to the driver's qualifications, the weight of the qualification adaptation coefficient may be relatively high. Through weighted calculation, the two coefficients are combined to obtain the first freight driver matching coefficient, which comprehensively reflects the overall matching degree of the first driver with the first freight order in terms of qualifications and time.

[0027] Step S236, determining whether the first freight driver registration coefficient is greater than / equal to the freight driver registration threshold. Specifically, determining whether the first freight driver registration coefficient is greater than / equal to the freight driver registration threshold, comparing the first freight driver registration coefficient with a preset freight driver registration threshold, the threshold is determined based on the platform's overall requirements and standards for freight drivers, and represents the minimum degree of matching for a driver to be considered to meet the order requirements.

[0028] Step S237: If the first freight driver registration coefficient is greater than / equal to the freight driver registration threshold, the first freight driver and the first freight driver registration coefficient are added to the first freight registration driver resource pool. Specifically, if the first freight driver registration coefficient is greater than / equal to the freight driver registration threshold, the first freight driver and the first freight driver registration coefficient are added to the first freight registration driver resource pool. If the first driver's registration coefficient meets the threshold requirement, it means that the driver is more suitable for the first freight order in terms of qualifications and time. The first driver's information and its registration coefficient are added to the first freight registration driver resource pool so that it can be used as one of the optional resources in subsequent order allocation.

[0029] Step S238, based on the freight driver registration weight condition and the freight driver registration threshold, the freight driver resource pool is continued to be registered in combination with the first freight order to obtain the first freight registered driver resource pool. Specifically, based on the freight driver registration weight condition and the freight driver registration threshold, the freight driver resource pool is continued to be registered in combination with the first freight order to obtain the first freight registered driver resource pool, after completing the evaluation and processing of the first driver, continue to register other drivers in the freight driver resource pool according to the same process and standards, and repeat the above steps until all drivers in the resource pool are evaluated, and finally obtain the complete first freight registered driver resource pool, which contains all suitable freight driver information that meets the requirements of the first freight order.

[0030] In a possible implementation, based on the freight vehicle registration threshold, the freight vehicle resource pool is registered according to the first freight order to generate a first freight registered vehicle resource pool, and step S240 further includes step S241, performing freight vehicle demand analysis based on the first freight order to generate first order vehicle qualification requirements and first order vehicle time requirements. Specifically, the freight vehicle demand analysis is performed based on the first freight order to generate the first order vehicle qualification requirements and first order vehicle time requirements, firstly analyzing the cargo characteristics in the first freight order, such as the volume, weight, and whether the cargo is a special item (such as fragile goods, dangerous goods, etc.), determining the qualifications that the required freight vehicles should have, such as load requirements, whether special protective equipment is required, etc., thereby generating the first order vehicle qualification requirements, considering the transportation time requirements and delivery deadline of the order, determining the time requirements for the freight vehicles, such as whether the vehicle can be used within a specific time period, and generating the first order vehicle time requirements.

[0031] Step S242, connect to the freight vehicle resource pool, and read the first vehicle qualification information and the first vehicle freight schedule of the first freight vehicle. Specifically, connect to the freight vehicle resource pool, read the first vehicle qualification information and the first vehicle freight schedule of the first freight vehicle, and after establishing a connection with the freight vehicle resource pool, the detailed information of each freight vehicle stored therein can be obtained. For the first freight vehicle, read its qualification information, including vehicle type, load capacity, special function configuration (such as refrigeration function, lifting equipment, etc.), vehicle age, etc., read the first vehicle freight schedule, understand the scheduled freight task time, maintenance time, etc. of the vehicle, so as to evaluate its time availability.

[0032] Step S243, based on the first order vehicle qualification requirements and the first vehicle qualification information, an adaptation evaluation is performed to obtain the first vehicle qualification adaptation coefficient. Specifically, based on the first order vehicle qualification requirements and the first vehicle qualification information, an adaptation evaluation is performed to obtain the first vehicle qualification adaptation coefficient, and the vehicle qualification requirements required by the first freight order are compared with the actual qualification information of the first vehicle. For example, if the order requires the vehicle load capacity to be at least 10 tons, and the load capacity of the first vehicle is 12 tons, then the degree of adaptation in the load capacity dimension is relatively high. A quantitative evaluation is performed on each qualification dimension, and finally the first vehicle qualification adaptation coefficient is obtained, which reflects the degree of matching of the first vehicle with the first freight order in terms of qualifications.

[0033] Step S244, based on the time requirement of the first order vehicle and the first vehicle freight schedule, an adaptation evaluation is performed to obtain the first vehicle freight time adaptation coefficient. Specifically, based on the time requirement of the first order vehicle and the first vehicle freight schedule, an adaptation evaluation is performed to obtain the first vehicle freight time adaptation coefficient, and the time requirement of the first freight order for the vehicle is compared with the freight schedule of the first vehicle. If the order requires the vehicle to be available within the next two days, and the first vehicle freight schedule is checked to find that there are no scheduled freight tasks within these two days and it is not during the maintenance time, then the adaptation degree in the time dimension is high, and the time adaptation degree is quantified to obtain the first vehicle freight time adaptation coefficient, which reflects the degree to which the first vehicle meets the requirements of the first freight order in terms of time.

[0034] Step S245, weighted calculation is performed on the first vehicle qualification adaptation coefficient and the first vehicle freight time adaptation coefficient based on the freight vehicle registration weight condition to obtain the first freight vehicle registration coefficient. Specifically, weighted calculation is performed on the first vehicle qualification adaptation coefficient and the first vehicle freight time adaptation coefficient based on the freight vehicle registration weight condition to obtain the first freight vehicle registration coefficient. The weights of the vehicle qualification adaptation coefficient and the freight time adaptation coefficient are determined according to the business focus and strategy of the freight platform. For example, if the platform pays more attention to the vehicle availability time, the weight of the freight time adaptation coefficient may be relatively high. Through weighted calculation, the two coefficients are combined to obtain the first freight vehicle registration coefficient, which comprehensively reflects the overall matching degree of the first vehicle with the first freight order in terms of qualification and time.

[0035] Step S246, determining whether the first freight vehicle registration coefficient is greater than / equal to the freight vehicle registration threshold. Specifically, determining whether the first freight vehicle registration coefficient is greater than / equal to the freight vehicle registration threshold, comparing the registration coefficient of the first freight vehicle with a preset freight vehicle registration threshold, the threshold is determined based on the platform's overall requirements and standards for freight vehicles, and represents the minimum degree of matching for a vehicle to be considered to meet the order requirements.

[0036] Step S247, if the first freight vehicle registration coefficient is greater than / equal to the freight vehicle registration threshold, the first freight vehicle and the first freight vehicle registration coefficient are added to the first freight registration vehicle resource pool. Specifically, if the first freight vehicle registration coefficient is greater than / equal to the freight vehicle registration threshold, the first freight vehicle and the first freight vehicle registration coefficient are added to the first freight registration vehicle resource pool. If the registration coefficient of the first vehicle meets the threshold requirement, it means that the vehicle is more suitable for the first freight order in terms of qualifications and time. The information of the first vehicle and its registration coefficient are added to the first freight registration vehicle resource pool so that it can be used as one of the optional resources in the subsequent order allocation.

[0037] Step S248, based on the freight vehicle registration weight condition and the freight vehicle registration threshold, the freight vehicle resource pool is continued to be registered in combination with the first freight order to obtain the first freight registered vehicle resource pool. Specifically, based on the freight vehicle registration weight condition and the freight vehicle registration threshold, the freight vehicle resource pool is continued to be registered in combination with the first freight order to obtain the first freight registered vehicle resource pool, and after completing the evaluation and processing of the first vehicle, the other vehicles in the freight vehicle resource pool are continued to be registered according to the same process and standards, and the above steps are repeated until all vehicles in the resource pool are evaluated, and finally a complete first freight registered vehicle resource pool is obtained, which contains all suitable freight vehicle information that meets the requirements of the first freight order.

[0038] Step S300, based on the first freight order and the first freight matching resource list, according to the freight order allocation model, output the first freight order allocation result. Specifically, there is a first freight order containing detailed information of the goods, and a first freight matching resource list established through the previous steps, which contains freight drivers and vehicle resource information that meet the order requirements. Then, the freight order allocation model is introduced, considering multiple factors such as transportation cost, transportation time, resource availability and reliability. After the first freight order and the first freight matching resource list are input into the model, the model extracts the key information of the order and evaluates the characteristics and capabilities of the resources, and allocates them by comprehensively considering various factors, such as calculating transportation cost, predicting transportation time, ensuring resource availability and reliability, etc. Finally, the first freight order allocation result is output, and the specific freight driver and vehicle combination, transportation route, estimated transportation time and cost, etc. are clarified as the basis for subsequent evaluation and optimization.

[0039] In a possible implementation, based on the first freight order and the first freight registration resource list, according to the freight order allocation model, the first freight order allocation result is output, and step S300 further includes step S310, according to the freight platform, loading a freight order sample set, a freight registration resource list sample set, and a freight order allocation result sample set. Specifically, according to the freight platform, the freight order sample set, the freight registration resource list sample set, and the freight order allocation result sample set are loaded. The freight platform, as a source of data and an application scenario, provides a basis for building a freight order allocation model. First, historical data is extracted from the freight platform and organized into different sample sets. The freight order sample set contains various types of freight order information, such as cargo type, quantity, origin, destination, transportation time requirements, etc. The freight registration resource list sample set includes freight resource information that has been screened in the past and meets the requirements of different freight orders, such as the characteristics and availability of freight drivers and freight vehicles, etc. The freight order allocation result sample set records the actual allocation results made for different freight orders in history, including the allocated freight drivers, freight vehicles, transportation routes, etc.

[0040] Step S320, supervised training is performed on multiple large models according to the freight order sample set, the freight registration resource list sample set and the freight order allocation result sample set, so as to generate multiple freight order distributors that meet the freight order allocation precision constraint. Specifically, supervised training is performed on multiple large models according to the freight order sample set, the freight registration resource list sample set and the freight order allocation result sample set, so as to generate multiple freight order distributors that meet the freight order allocation precision constraint. The three sample sets are used to train multiple large models, and the large model can be an algorithm model with strong learning ability such as a deep learning model. In the training process, the order information in the freight order sample set and the resource information in the freight registration resource list sample set are used as inputs, and the actual allocation result in the freight order allocation result sample set is used as the target output. By continuously adjusting the parameters of the model, the model can learn how to reasonably allocate according to the order and resource information. At the same time, the freight order allocation precision constraint is set, such as requiring the accuracy of the allocation result to reach a certain level, the prediction error of the transportation time and cost to be within a certain range, etc. Only the model that meets these constraints can be retained as the freight order distributor.

[0041] Step S330, using the output data sets of the multiple freight order distributors as input information, and the freight order distribution result sample set as output information, to train a freight order distribution aggregator. Specifically, using the output data sets of the multiple freight order distributors as input information, and the freight order distribution result sample set as output information, to train a freight order distribution aggregator. Multiple freight order distributors may produce different results when distributing new orders and resources. In order to integrate these results and improve the accuracy and stability of distribution, a freight order distribution aggregator is introduced, and the output data sets of the multiple freight order distributors are used as input information of the aggregator. These output data sets contain different distribution plans of each distributor for the same order and resource. Using the freight order distribution result sample set as output information, the aggregator is trained so that it can learn how to integrate the results of multiple distributors to obtain the final distribution plan.

[0042] Step S340, merging the multiple freight order distributors as parallel independent nodes to generate a freight order distribution processing layer. Specifically, the multiple freight order distributors are merged as parallel independent nodes to generate a freight order distribution processing layer, and the multiple freight order distributors are regarded as parallel independent nodes, and they are merged to form a freight order distribution processing layer. When processing a new freight order, this processing layer can use multiple distributors to analyze and distribute orders and resources at the same time, give full play to the advantages of each distributor, and improve the efficiency and accuracy of distribution.

[0043] Step S350, merging the freight order allocation processing layer with the input layer of the freight order allocation aggregator to generate the freight order allocation model. Specifically, merging the freight order allocation processing layer with the input layer of the freight order allocation aggregator to generate the freight order allocation model, integrating the freight order allocation processing layer and the freight order allocation aggregator, the freight order allocation processing layer is responsible for generating multiple allocation plans, and the freight order allocation aggregator is responsible for aggregating and optimizing these plans, the two are combined to form a complete freight order allocation model, the model can quickly and accurately output the optimal allocation result according to the new freight order and resource information.

[0044] Step S400, based on the first freight order allocation result, according to the order allocation fitness function, calculate the first order allocation fitness. Specifically, based on the existing first freight order allocation result, an order allocation fitness function is introduced, and the function comprehensively considers factors such as transportation time, transportation cost, and cargo safety. Then, according to the allocation result, relevant parameters such as expected transportation time and expected transportation cost are extracted, and at the same time, information such as expected transportation time and expected transportation cost is obtained from the first freight order. Taking the specific fitness function as an example, First, the freight timeliness evaluation coefficient, freight cost control evaluation coefficient and freight accident prediction probability are calculated and normalized to obtain g(FEX), g(FCX), and g(FLX). Then, the weights α, β, and γ are substituted into the function calculation to finally obtain the first order allocation fitness, which is used to evaluate the current allocation results and provide a basis for subsequent judgment and optimization.

[0045] In a possible implementation, based on the first freight order allocation result, according to an order allocation fitness function, the first order allocation fitness is calculated, and step S400 further includes step S410, where the order allocation fitness function is: ; Among them, OAF represents the fitness of order allocation, exp represents the exponential function with the natural constant e as the base, g(FEX) represents the normalized freight time evaluation coefficient, g(FCX) represents the normalized freight cost control evaluation coefficient, g(FLX) represents the normalized freight accident prediction probability, α, β, γ represent the order allocation fitness calculation weights. Specifically, the order allocation fitness function It is used to evaluate the fitness of freight order allocation, where OAF represents the fitness of order allocation, exp is an exponential function with the natural constant e as the base, g(FEX) is the normalized freight time evaluation coefficient, which is calculated and normalized by the difference between the actual transportation time and the expected transportation time. If the actual transportation time is much shorter than the expected transportation time, the coefficient is higher. g(FCX) is the normalized freight cost control evaluation coefficient, which is calculated and normalized based on factors such as the actual transportation cost. When the actual transportation cost is low, it plays a positive role in the function. g(FLX) is the normalized freight accident The prediction probability is calculated and normalized through the freight accident prediction model. A low accident prediction probability can reduce the negative impact on fitness. α, β, and γ are the fitness calculation weights for order allocation, which are determined according to the business objectives, customer needs, and market environment of the freight platform. They are used to adjust the importance of different factors in the fitness calculation. The calculation process is to first calculate g(FEX), g(FCX), and g(FLX) respectively, and then substitute them and the weights into the function to calculate the value in the brackets. Finally, the natural index is taken to obtain OAF, so as to quantitatively evaluate different freight order allocation schemes to select the optimal scheme.

[0046] In a possible implementation, based on the first freight order allocation result, according to the order allocation fitness function, the first order allocation fitness is calculated, and step S400 further includes step S420, obtaining the expected freight time and expected freight cost of the first freight order. Specifically, the expected freight time and expected freight cost of the first freight order are obtained. The first freight order usually specifies the customer's time requirements and cost budget for the transportation of goods. This information is an important reference standard for evaluating the order allocation result. By carefully analyzing the content of the first freight order, the customer's expected freight time, that is, the expected time range for the transportation of goods from the origin to the destination, can be determined, and the customer's expected freight cost, including the transportation fee, possible additional fees and other budget constraints, can be specified.

[0047] Step S430: forecasting the freight time based on the first freight order allocation result to obtain the first forecasted freight time, and evaluating the freight time efficiency of the first forecasted freight time according to the expected freight time to generate the first freight time efficiency evaluation coefficient. Specifically, forecasting the freight time based on the first freight order allocation result to obtain the first forecasted freight time, and evaluating the freight time efficiency of the first forecasted freight time according to the expected freight time to generate the first freight time efficiency evaluation coefficient. According to the freight resources (such as freight drivers, freight vehicles, transportation routes, etc.) determined in the first freight order allocation result, the freight time is forecasted using relevant algorithms and models. The forecasting method may include considering factors such as transportation distance, vehicle speed, and traffic conditions. After obtaining the first forecasted freight time, it is compared with the expected freight time. If the forecasted freight time is close to or less than the expected freight time, it means that the freight time efficiency is high; conversely, if the forecasted freight time is much greater than the expected freight time, it means that the freight time efficiency is low. The comparison result is converted into a numerical value, namely the first freight time efficiency evaluation coefficient, which reflects the performance of the order allocation result in terms of freight time.

[0048] Step S440, based on the first freight order allocation result, a freight cost forecast is made to obtain a first forecast freight cost, and the first forecast freight cost is evaluated for freight cost control in combination with the expected freight cost to obtain a first freight cost control evaluation coefficient. Specifically, based on the first freight order allocation result, a freight cost forecast is made to obtain a first forecast freight cost, and the first forecast freight cost is evaluated for freight cost control in combination with the expected freight cost to obtain a first freight cost control evaluation coefficient. According to the first freight order allocation result, various cost factors in the transportation process, such as fuel costs, driver wages, tolls, etc., are considered to make a freight cost forecast to obtain a first forecast freight cost, and the first forecast freight cost is compared with the expected freight cost. If the forecast freight cost is within the expected freight cost range or lower than the expected freight cost, it means that the freight cost control is good; if the forecast freight cost is much higher than the expected freight cost, it means that the cost control is poor, and the comparison result is converted into a first freight cost control evaluation coefficient, which reflects the control effect of the order allocation result on the freight cost.

[0049] Step S450, based on the first freight order allocation result, according to the freight accident prediction model, obtain the first freight accident prediction probability. Specifically, based on the first freight order allocation result, according to the freight accident prediction model, obtain the first freight accident prediction probability. Freight accidents may have a serious impact on the transportation of goods, so they need to be predicted. According to the freight resources and transportation conditions determined in the first freight order allocation result, the freight accident prediction model is used for analysis. The model will consider the experience and safety record of the freight driver, the maintenance status of the vehicle, the safety of the transportation route and other factors to obtain the first freight accident prediction probability, that is, the possibility of an accident occurring during the transportation of the goods.

[0050] Step S460, based on the first freight timeliness evaluation coefficient, the first freight cost control evaluation coefficient and the first freight accident prediction probability, normalization processing is performed, and the first order allocation fitness is generated in combination with the order allocation fitness function. Specifically, based on the first freight timeliness evaluation coefficient, the first freight cost control evaluation coefficient and the first freight accident prediction probability, normalization processing is performed, and the first order allocation fitness is generated in combination with the order allocation fitness function. The first freight timeliness evaluation coefficient, the first freight cost control evaluation coefficient and the first freight accident prediction probability are normalized and converted into values ​​within a specific range so as to perform comprehensive calculations in the order allocation fitness function. The normalization method can ensure that the weights of different factors in the calculation are more reasonable. Substitute the normalized coefficients into the order allocation fitness function, such as , where g(FEX) is the normalized freight time evaluation coefficient, g(FCX) is the normalized freight cost control evaluation coefficient, g(FLX) is the normalized freight accident prediction probability, α, β, γ are the order allocation fitness calculation weights, and the first order allocation fitness is finally generated through function calculation. This fitness value comprehensively reflects the overall performance of the first freight order allocation result in terms of freight time, freight cost, and freight safety, and can be used to evaluate the quality of the order allocation results.

[0051] In one possible implementation, based on the first freight order allocation result and according to the freight accident prediction model, a first freight accident prediction probability is obtained, and step S450 further includes step S451, obtaining freight accident prediction-related factors, wherein the freight accident prediction-related factors include freight driver factors, freight vehicle factors, freight environment factors, and cargo factors. Specifically, the prediction of freight accidents requires comprehensive consideration of multiple factors, wherein freight driver factors include the driver's driving experience, such as total driving time, familiarity with the current transportation route, historical accident records, etc.; the driver's health status and fatigue level. Long-term driving may lead to fatigue and increase the risk of accidents, reflecting the impact of the driver's ability and status on freight safety. Freight vehicle factors include the maintenance status of the vehicle, that is, whether the vehicle has been regularly maintained and the status of key components such as tires and brakes; the age of the vehicle. Vehicles with longer service life may have a higher risk of failure. The condition of the vehicle is directly related to the safety during transportation. Freight environmental factors include weather conditions. Conditions, such as bad weather (rain, snow, fog) will increase the possibility of traffic accidents; road conditions, including the complexity of the road and whether there are high-risk sections (such as sharp turns, mountain roads, etc.); traffic conditions, such as traffic congestion or transportation during peak hours may increase the probability of accidents. Environmental factors are external conditions that have an impact on freight safety that cannot be ignored. Cargo factors include cargo type, whether the transported cargo is dangerous goods or cargo that requires special handling (such as flammables, chemicals, etc.); cargo weight and loading method. Excessive or uneven loading of cargo may affect the balance and braking performance of the vehicle, thereby increasing the risk of accidents. The characteristics and loading conditions of the cargo will also have an important impact on transportation safety.

[0052] Step S452, collect the associated parameters of the first freight order allocation result according to the freight accident prediction associated factors to obtain the first freight accident prediction associated data set. Specifically, collect the associated parameters of the first freight order allocation result according to the freight accident prediction associated factors to obtain the first freight accident prediction associated data set. After clarifying the prediction associated factors, perform specific data collection work on the first freight order allocation result. For example, for driver factors, query the driving experience information, current health status and fatigue level of the assigned driver; for vehicle factors, check the maintenance record, service life and key component status of the assigned vehicle; for environmental factors, obtain the weather forecast, road conditions and traffic flow information on the transportation route; for cargo factors, determine the type, weight and loading method of the cargo. Collect these parameters related to freight accidents to form the first freight accident prediction associated data set.

[0053] Step S453, based on the first freight accident prediction associated data set, according to the freight accident prediction model, obtain the first freight accident prediction probability. Specifically, based on the first freight accident prediction associated data set, according to the freight accident prediction model, obtain the first freight accident prediction probability, use the collected first freight accident prediction associated data set, and input it into the freight accident prediction model. The model is trained based on historical data, and can comprehensively evaluate the possibility of freight accidents according to various related factors, analyze and process each factor in the data set, and calculate the first freight accident prediction probability by combining the weights and mutual relationships of different factors. The probability value reflects the possibility of an accident occurring during the freight transportation process under the current first freight order allocation result, and provides an important basis for evaluating the rationality and safety of order allocation.

[0054] Step S500, determine whether the first order allocation fitness satisfies the order allocation fitness constraint. Specifically, the order allocation fitness is a numerical value that comprehensively reflects the performance of the freight order allocation results in multiple aspects, and the order allocation fitness constraint is a pre-set standard or range. After obtaining the first order allocation fitness, it is compared with the order allocation fitness constraint. If the constraint is a numerical range, it is determined whether the fitness falls within the range. If the constraint is a plurality of specific indicator requirements, each indicator corresponding to the fitness is compared with the constraint indicator. If the first order allocation fitness satisfies the constraint conditions, it means that the current order allocation plan meets the basic requirements of the freight platform; if not, it means that it fails to meet the standards in some aspects. This judgment provides a decision-making basis for subsequent order allocation optimization.

[0055] Step S600, if the first order allocation fitness does not meet the order allocation fitness constraint, generate the first order allocation optimization instruction. Specifically, when the first order allocation fitness does not meet the order allocation fitness constraint, it means that the current freight order allocation plan does not meet the expected requirements of the freight platform. At this time, the first order allocation optimization instruction needs to be generated. First, analyze the gap between the fitness and the constraint to determine the optimization direction, consider the relationship between various factors, such as shortening the transportation time may increase the cost, etc., comprehensively determine the balanced optimization direction, and generate specific instructions according to the optimization direction, including changing drivers or vehicles, adjusting routes, rescheduling time, etc., which can be based on preset rule algorithms or combined with manual experience. Finally, the instructions are passed to relevant system modules or personnel for execution, and the freight order allocation plan is adjusted and improved, such as the system automatically realigns the allocation resources or manually intervenes to adjust the transportation plan.

[0056] Step S700, based on the first order allocation optimization instruction, the first freight order allocation result is optimized and allocated according to the order allocation fitness function, the order allocation fitness constraint, the first freight registration resource list and the first freight order, to generate the first freight order allocation optimization result. Specifically, based on the first order allocation optimization instruction, it is clear that the current first freight order allocation result needs to be optimized, and factors such as the order allocation fitness function, the order allocation fitness constraint, the first freight registration resource list and the first freight order are considered, and the problems existing in the current allocation result, such as long transportation time or high cost, are analyzed, and appropriate freight drivers and vehicle resources are reselected according to the resource list, and the transportation route is optimized. The fitness of the adjusted allocation plan is calculated and the constraints are checked. After multiple adjustments, the plan with the highest fitness and meeting the constraints is selected as the first freight order allocation optimization result to meet the freight order needs and platform requirements to the greatest extent.

[0057] In one possible implementation, Figure 2 As shown, based on the first order allocation optimization instruction, the first freight order allocation result is optimized and allocated according to the order allocation fitness function, the order allocation fitness constraint, the first freight registration resource list and the first freight order, and the first freight order allocation optimization result is generated. Step S700 further includes step S710, taking the first freight order as the order allocation target, adjusting the first freight order allocation result in combination with the first freight registration resource list, and establishing a freight order allocation adjustment space. Specifically, taking the first freight order as the order allocation target means that the entire plan is carried out around meeting the needs of this specific order. In combination with the first freight registration resource list, which is the resource information such as freight drivers and freight vehicles that meet the requirements of the first freight order obtained through preliminary screening, the first freight order allocation result is adjusted, which may include replacing different drivers and vehicle combinations, adjusting transportation routes, time arrangements, etc. Through various possible adjustment methods, multiple freight order allocation adjustment decisions are formed, and these decisions together constitute the freight order allocation adjustment space, which provides multiple possibilities for the subsequent optimization process.

[0058] Step S720, based on the freight order allocation adjustment space, extract the first freight order allocation adjustment decision. Specifically, based on the freight order allocation adjustment space, extract one from a plurality of adjustment decisions as the first freight order allocation adjustment decision, which can be randomly extracted or selected according to certain rules, such as giving priority to certain specific types of adjustment decisions, or selecting according to the experience of historical data.

[0059] Step S730, based on the first freight order allocation adjustment decision, calculate the first adjustment decision order allocation fitness according to the order allocation fitness function. Specifically, based on the first freight order allocation adjustment decision, the order allocation fitness function is used for calculation, and the function is calculated in the same way as the first order allocation fitness, taking into account multiple factors such as the freight timeliness evaluation coefficient, the freight cost control evaluation coefficient, and the freight accident prediction probability, and extracting relevant parameters such as the expected transportation time and transportation cost according to the specific content of the adjustment decision, and combining the expected transportation time, expected transportation cost and other information in the first freight order, and substituting them into the function for calculation, and finally obtaining the first adjustment decision order allocation fitness, which reflects the pros and cons of the current adjustment decision after comprehensive consideration of multiple aspects.

[0060] Step S740, judging whether the first adjustment decision order allocation fitness satisfies the order allocation fitness constraint. Specifically, judging whether the first adjustment decision order allocation fitness satisfies the order allocation fitness constraint, the calculated fitness value is compared with a preset constraint condition, and the constraint condition is determined according to the requirements and objectives of the freight business, such as restrictions on transportation time, cost, accident risk, etc.

[0061] Step S750: If the first adjustment decision order allocation fitness satisfies the order allocation fitness constraint, the first freight order allocation adjustment decision is added to the first freight order allocation optimization result. Specifically, if the first adjustment decision order allocation fitness satisfies the order allocation fitness constraint, it means that the adjustment decision is a better solution, and the first freight order allocation adjustment decision is added to the first freight order allocation optimization result as part of the final optimization result.

[0062] Step S760: If the fitness of the first adjustment decision order allocation does not satisfy the fitness constraint of the order allocation, the first freight order allocation adjustment decision is eliminated, and the freight order allocation adjustment space is continuously optimized according to the fitness function of the order allocation and the fitness constraint of the order allocation, until the first freight order allocation optimization result is generated. Specifically, if the fitness of the first adjustment decision order allocation does not satisfy the fitness constraint of the order allocation, it means that the adjustment decision is not feasible, and the decision is eliminated. Then, according to the fitness function of the order allocation and the fitness constraint of the order allocation, the freight order allocation adjustment space is continuously optimized, and a new adjustment decision is extracted from the adjustment space again, and the process of calculating the fitness and judging the constraint is repeated, and the cycle is continuously repeated until an adjustment decision that satisfies the constraint condition is found or a certain stop condition is reached. If a decision that satisfies the condition is still not found after a certain number of iterations, the first freight order allocation optimization result is finally generated. This result is the order allocation plan that best meets the requirements and goals of the freight business after continuous adjustment and optimization.

[0063] The embodiment of the present application adopts a method of connecting to a freight platform to read a first freight order, establishing a first freight matching resource list based on freight resource matching constraints, outputting a distribution result through a freight order distribution model in combination with the first freight order and the list, calculating fitness according to an order distribution fitness function, judging whether the constraints are met, and generating optimization instructions if not, optimizing the first freight order distribution result according to the instructions and related content, and generating an optimization result, thereby achieving a technical effect of improving the operational efficiency of the freight platform by achieving accurate matching of freight resources, scientifically and rationally allocating orders, and ensuring high fitness.

[0064] The above specific implementation manner does not constitute a limitation to the protection scope of the present application. It should be understood by those skilled in the art that various modifications, combinations and substitutions can be made according to design requirements and other factors. Any modifications, equivalent substitutions and improvements made within the spirit and principles of the present application should be included in the protection scope of the present application. In some cases, the actions or steps recorded in the present application can be performed in an order different from that in the embodiment and can still achieve the desired results. In addition, the process depicted in the accompanying drawings does not necessarily require the specific order or continuous order shown to achieve the desired results. In some embodiments, multitasking and parallel processing are also possible or may be advantageous.

Claims

1. A method for optimizing intelligent order allocation for freight platforms based on a deep learning big model, characterized in that: The method comprises: Connecting to the freight order management module of the freight platform to read the first freight order; Based on the freight resource registration constraint, performing resource registration on the freight resource pool of the freight platform according to the first freight order, and establishing a first freight registration resource list; Based on the first freight order and the first freight registration resource list, outputting a first freight order allocation result according to a freight order allocation model; Based on the first freight order allocation result, the first order allocation fitness is calculated according to the order allocation fitness function, wherein the order allocation fitness function is: OAF=exp [ (α*g(FEX)+β*g(FCX)-γ*g(FLX) ] ; Among them, OAF represents the fitness of order allocation, exp represents the exponential function with the natural constant e as the base, g(FEX) represents the normalized freight time evaluation coefficient, g(FCX) represents the normalized freight cost control evaluation coefficient, g(FLX) represents the normalized freight accident prediction probability, α, β, γ represent the order allocation fitness calculation weights; Determining whether the first order allocation fitness satisfies the order allocation fitness constraint; If the first order allocation fitness does not satisfy the order allocation fitness constraint, generating a first order allocation optimization instruction; Based on the first order allocation optimization instruction, optimizing and allocating the first freight order allocation result according to the order allocation fitness function, the order allocation fitness constraint, the first freight registration resource list and the first freight order, to generate a first freight order allocation optimization result; Based on the freight resource registration constraint, the freight resource pool of the freight platform is registered with the first freight order to establish a first freight registration resource list, including: The freight resource registration constraints include a freight driver registration threshold and a freight vehicle registration threshold; The freight resource pool includes a freight driver resource pool and a freight vehicle resource pool; Based on the freight driver registration threshold, the freight driver resource pool is registered according to the first freight order to generate a first freight registered driver resource pool; Based on the freight vehicle registration threshold, registering the freight vehicle resource pool according to the first freight order to generate a first freight registered vehicle resource pool; Arranging the first freight registration driver resource pool and the first freight registration vehicle resource pool to generate the first freight registration resource list; The steps of constructing the freight order allocation model include: According to the freight platform, a freight order sample set, a freight registration resource list sample set and a freight order allocation result sample set are loaded; According to the freight order sample set, the freight registration resource list sample set and the freight order allocation result sample set, multiple deep learning large models are supervised trained to generate multiple freight order allocators that meet the freight order allocation accuracy constraint; Using the output data sets of the plurality of freight order allocators as input information and the freight order allocation result sample set as output information, training a freight order allocation amalgamator; The multiple freight order distributors are merged as parallel independent nodes to generate a freight order distribution processing layer; the freight order distribution processing layer is merged with the input layer of the freight order distribution aggregator to generate the freight order distribution model, wherein the freight order distribution processing layer is responsible for generating multiple distribution plans, and the freight order distribution aggregator is responsible for fusing and optimizing the multiple distribution plans.

2. The method according to claim 1, characterized in that Based on the freight driver registration threshold, the freight driver resource pool is registered according to the first freight order to generate a first freight registered driver resource pool, including: Analyze the demand for freight drivers based on the first freight order, and generate the qualification requirements and time requirements of the drivers for the first order; Connecting to the freight driver resource pool, reading the first driver qualification information and the first driver freight schedule of the first freight driver; Performing an adaptation evaluation based on the first order driver qualification requirement and the first driver qualification information to obtain a first driver qualification adaptation coefficient; Performing an adaptation evaluation based on the time requirement of the driver for the first order and the freight time schedule of the first driver to obtain a freight time adaptation coefficient of the first driver; Performing weighted calculation on the first driver qualification adaptation coefficient and the first driver freight time adaptation coefficient based on the freight driver matching weight condition to obtain the first freight driver matching coefficient; Determining whether the first freight driver registration coefficient is greater than / equal to the freight driver registration threshold; If the first freight driver registration coefficient is greater than / equal to the freight driver registration threshold, adding the first freight driver and the first freight driver registration coefficient to the first freight registration driver resource pool; Based on the freight driver registration weight condition and the freight driver registration threshold, the freight driver resource pool is continuously registered in combination with the first freight order to obtain the first freight registered driver resource pool.

3. The method according to claim 1, characterized in that Based on the freight vehicle registration threshold, the freight vehicle resource pool is registered according to the first freight order to generate a first freight registered vehicle resource pool, including: Analyze the demand for freight vehicles based on the first freight order to generate the first order vehicle qualification requirements and the first order vehicle time requirements; Connecting to the freight vehicle resource pool, reading first vehicle qualification information of a first freight vehicle and a freight schedule of the first vehicle; Performing an adaptation evaluation based on the first order vehicle qualification requirement and the first vehicle qualification information to obtain a first vehicle qualification adaptation coefficient; Performing an adaptation evaluation based on the first order vehicle time requirement and the first vehicle freight time schedule to obtain a first vehicle freight time adaptation coefficient; Performing weighted calculation on the first vehicle qualification adaptation coefficient and the first vehicle freight time adaptation coefficient based on the freight vehicle registration weight condition to obtain a first freight vehicle registration coefficient; Determining whether the first freight vehicle registration coefficient is greater than / equal to the freight vehicle registration threshold; If the first freight vehicle registration coefficient is greater than / equal to the freight vehicle registration threshold, adding the first freight vehicle and the first freight vehicle registration coefficient to the first freight registration vehicle resource pool; Based on the freight vehicle registration weight condition and the freight vehicle registration threshold, the freight vehicle resource pool is continuously registered in combination with the first freight order to obtain the first freight registered vehicle resource pool.

4. The method according to claim 1, characterized in that Based on the first order allocation optimization instruction, the first freight order allocation result is optimized and allocated according to the order allocation fitness function, the order allocation fitness constraint, the first freight registration resource list and the first freight order to generate a first freight order allocation optimization result, including: Taking the first freight order as the order allocation target, adjusting the first freight order allocation result in combination with the first freight registration resource list, and establishing a freight order allocation adjustment space; extracting a first freight order allocation adjustment decision based on the freight order allocation adjustment space; Based on the first freight order allocation adjustment decision, and according to the order allocation fitness function, calculating the first adjustment decision order allocation fitness; Determining whether the first adjustment decision order allocation fitness satisfies the order allocation fitness constraint; If the first adjustment decision order allocation fitness satisfies the order allocation fitness constraint, adding the first freight order allocation adjustment decision to the first freight order allocation optimization result; If the first adjustment decision order allocation fitness does not satisfy the order allocation fitness constraint, the first freight order allocation adjustment decision is eliminated, and the freight order allocation adjustment space is continued to be optimized according to the order allocation fitness function and the order allocation fitness constraint until the first freight order allocation optimization result is generated.

5. The method according to claim 1, characterized in that Based on the first freight order allocation result, calculating the first order allocation fitness according to the order allocation fitness function includes: Obtaining an expected shipping time and an expected shipping cost for the first shipping order; Predicting the freight time based on the first freight order allocation result to obtain a first predicted freight time, and evaluating the first predicted freight time according to the expected freight time to generate a first freight time evaluation coefficient; Performing a freight cost forecast based on the first freight order allocation result to obtain a first forecasted freight cost, and performing a freight cost control evaluation on the first forecasted freight cost in combination with the expected freight cost to obtain a first freight cost control evaluation coefficient; Based on the first freight order allocation result, and according to the freight accident prediction model, a first freight accident prediction probability is obtained; The first order allocation fitness is generated by performing normalization processing based on the first freight timeliness evaluation coefficient, the first freight cost control evaluation coefficient and the first freight accident prediction probability, and combining the order allocation fitness function.

6. The method according to claim 5, characterized in that Based on the first freight order allocation result and according to the freight accident prediction model, a first freight accident prediction probability is obtained, including: Obtaining freight accident prediction correlation factors, wherein the freight accident prediction correlation factors include freight driver factors, freight vehicle factors, freight environment factors and freight factors; Collecting correlation parameters of the first freight order allocation result according to the freight accident prediction correlation factors to obtain a first freight accident prediction correlation data set; Based on the first freight accident prediction associated data set and according to a freight accident prediction model, the first freight accident prediction probability is obtained.

Images