Waybill adaptive calculation method and system based on dynamic transportation price rule engine

Abstract

The present application relates to the technical field of waybill price calculation, and discloses a waybill adaptive calculation method and system based on a dynamic transportation price rule engine, the method comprising obtaining the user's transportation demand data and idle vehicle data; calculating the theoretical unit freight of each idle vehicle based on the dynamic transportation price rule, transportation demand data, and idle vehicle data; screening alternative vehicles and constructing an alternative vehicle set; outputting a transportation vehicle combination and the theoretical transportation cost of each transportation vehicle combination through iterative optimization; correcting the theoretical transportation cost based on historical waybill data to generate a probability distribution of the order price; generating a recommended price range based on the probability distribution of the order price; and performing feedback optimization on the generation of the recommended price range based on the order price selected by the user. The present application can provide users with an accurate waybill price reference by intelligently matching transportation resources.

Description

Technical Field

[0001] The present application relates to the technical field of waybill price calculation, and in particular to a waybill adaptive calculation method and system based on a dynamic transportation price rule engine. Background Art

[0002] Traditional logistics waybill pricing relies on manual experience and struggles to respond to market fluctuations in real time. Small and medium-sized shippers lack real-time price perception in the transportation market, and their self-set pricing can easily deviate from reasonable market ranges. This leads to low driver acceptance rates, high vehicle vacancy rates, and inefficient matching of shipper needs. Existing transportation cost calculations are often crude. Faced with diverse transportation needs, traditional solution generation models struggle to adapt quickly. They often rely on a single, fixed waybill calculation model, failing to provide personalized solutions.

[0003] There's a natural conflict between shipping costs and order acceptance rates: while lowering freight rates can attract shippers, it can also lead to an increase in driver rejection rates. Conversely, while higher freight rates increase drivers' willingness to accept orders, they weaken the platform's competitiveness. Existing systems fail to balance multiple objectives, such as order acceptance rate and delivery timeliness, resulting in significant discrepancies between theoretically calculated and actual order prices. Drivers' order acceptance decisions are influenced by both subjective and objective factors, and existing platforms lack the ability to mine drivers' historical behavior. Due to a lack of dynamic modeling of drivers' acceptance thresholds, platform-recommended fares often deviate from drivers' expectations, leading to a high rate of order matching failures.

[0004] For example, Chinese patent application publication number CN115587779A discloses a method for fault-tolerant estimation of freight transportation prices based on spatiotemporal dynamic information, the method comprising: matching historical waybill data based on acquired key transportation information, and determining whether the number of matched historical waybill data meets a first preset quantity threshold; if so, estimating the transportation price range corresponding to the vehicle to be queried based on the freight information of the historical waybill data; if not, further determining whether the number of matched historical waybill data meets a second preset quantity threshold; if so, estimating the transportation price range corresponding to the vehicle to be queried based on the freight information of the historical waybill data; if not, estimating the transportation price range based on the path information in the key transportation information of the vehicle to be queried, and adjusting the estimated transportation price range based on the waybill attributes in the key transportation information. This technical solution can significantly improve the accuracy, reliability, and computational efficiency of freight price estimation, but it still faces the problem raised in the background technology of this application: it adopts a relatively single and fixed waybill trial calculation model, which cannot provide users with personalized solutions.

[0005] The information disclosed in this background technology section is only intended to enhance the understanding of the overall background of the application and should not be regarded as an admission or any form of suggestion that the information constitutes the prior art already known to ordinary technicians in this field. Summary of the Invention

[0006] The technical problem to be solved by this application is to overcome the defects of the existing technology and provide a waybill adaptive calculation method and system based on a dynamic transportation price rule engine, which provides users with accurate waybill price references by intelligently matching transportation resources.

[0007] To solve the above technical problems, this application provides the following technical solutions:

[0008] In one aspect, the present application provides a waybill adaptive calculation method based on a dynamic transportation price rule engine, comprising the following steps:

[0009] Obtain users' transportation demand data and idle vehicle data;

[0010] Loading preset dynamic transportation pricing rules; calculating the theoretical unit freight for each idle vehicle based on the dynamic transportation pricing rules, transportation demand data, and idle vehicle data;

[0011] Based on the transportation demand data and the idle vehicle data, screening candidate vehicles and building a candidate vehicle set;

[0012] Based on the candidate vehicle set and the theoretical unit freight rate, outputting a transport vehicle combination and a theoretical transport cost of each transport vehicle combination through iterative optimization;

[0013] Obtain historical waybill data; modify the theoretical transportation cost based on the historical waybill data to generate a probability distribution of the order price;

[0014] generating a recommended price range based on the probability distribution of the order price and the theoretical transportation cost;

[0015] Based on the order price selected by the user, feedback optimization is performed on the generation of the recommended price range.

[0016] As a preferred solution of the adaptive calculation method for waybills based on the dynamic transportation price rule engine described in this application, the transportation demand data includes cargo type, cargo weight, starting point, destination, and deadline; the idle vehicle data includes a list of idle vehicles, the model, load capacity, current location, idle time, remaining available time, and driver credit score of each idle vehicle;

[0017] The theoretical unit freight is the freight for transporting unit weight of goods per unit distance when an idle vehicle executes a waybill issued by a user. The method for calculating the theoretical unit freight based on the dynamic transportation price rule includes:

[0018] Determining a transportation route and a delivery date based on the transportation demand data; wherein the transportation route is determined based on the starting point and the end point; and the delivery date is determined based on the cut-off time;

[0019] Call the weather API to obtain the weather conditions for the transportation route during the delivery date;

[0020] Based on the transportation route, delivery date, weather conditions, cargo type, model and load capacity of the idle vehicle, the theoretical unit freight for each idle vehicle when executing the waybill issued by the user is dynamically calculated.

[0021] As a preferred solution of the adaptive calculation method for shipping orders based on the dynamic transportation price rule engine described in this application, the iterative optimization output of the transport vehicle combination and the theoretical transportation cost of each transport vehicle combination specifically includes:

[0022] Generate at least m groups of alternative vehicle combinations iteratively through the genetic algorithm, and record the theoretical transportation cost of each alternative vehicle combination; m is a positive integer;

[0023] A non-dominated solution is screened from the alternative vehicle combinations as a transport vehicle combination; the transport vehicle combination includes a first combination and a second combination, wherein the first combination is the transport vehicle combination with the lowest theoretical transport cost, and the second combination is the transport vehicle combination with the highest order acceptance rate.

[0024] As a preferred solution of the adaptive calculation method of shipping orders based on the dynamic transportation price rule engine described in this application, the genetic algorithm specifically includes:

[0025] S100: Generate an idle vehicle combination for receiving orders; assign a task amount to each idle vehicle in each idle vehicle combination to obtain an initial population; the task amount is the cargo weight corresponding to the idle vehicle receiving the order;

[0026] S200: Calculate an optimization target value for each idle vehicle combination; record idle vehicle combinations whose optimization target values ​​are less than a preset optimization target threshold as candidate vehicle combinations; and record the number of candidate vehicle combinations;

[0027] S300: Generate the next generation population through selection operation, crossover operation and mutation operation;

[0028] S400: iteratively executing S200-S300 until the number of candidate vehicle combinations is greater than or equal to m, and then stopping the iteration;

[0029] S500: Outputting the alternative vehicle combinations and the theoretical transportation cost of each alternative vehicle combination;

[0030] In the genetic algorithm, the common constraints of each generation of population include: the sum of the loads of all idle vehicles in any idle vehicle combination is not less than the cargo weight. times, and not more than the weight of the goods times; among them, is the preset lower load limit multiple, It is a multiple of the preset upper load limit.

[0031] As a preferred solution of the adaptive calculation method for waybills based on the dynamic transportation price rule engine described in this application, the method for calculating the optimization target value of any idle vehicle combination is as follows:

[0032] Calculating the theoretical freight of each idle vehicle based on the task volume of each idle vehicle and the theoretical unit freight; calculating the sum of the theoretical freight of each idle vehicle in the idle vehicle combination to obtain the theoretical transportation cost of the idle vehicle combination;

[0033] Calculating the idle distance of each idle vehicle based on the current position of each idle vehicle;

[0034] Calculate the order acceptance cost factor of each idle vehicle based on the idle distance, idle time, and remaining available time of each idle vehicle; calculate the sum of the order acceptance cost factors of each idle vehicle in the idle vehicle combination to obtain the order acceptance cost factor of the idle vehicle combination;

[0035] The theoretical transportation cost of the idle vehicle combination is normalized and then weighted summed with the order acceptance cost factor of the idle vehicle combination to obtain the optimization target value of the idle vehicle combination.

[0036] As a preferred solution of the adaptive calculation method for waybills based on the dynamic transportation price rule engine described in this application, wherein: selecting a non-dominated solution from the alternative vehicle combinations as the transportation vehicle combination specifically includes:

[0037] Each alternative vehicle combination is considered a candidate solution. All candidate solutions form a solution space. The theoretical transportation cost and order acceptance cost factor corresponding to each candidate solution in the solution space are recorded. Non-dominated solutions in the solution space are iteratively screened, specifically including: if for any candidate solution A, there exists at least one candidate solution B such that the theoretical transportation cost and order acceptance cost factor of candidate solution B are both no greater than those of candidate solution A, and at least one of the theoretical transportation cost and order acceptance cost factors of candidate solution B is less than the corresponding item of candidate solution A, then candidate solution B dominates candidate solution A; otherwise, candidate solution A is a non-dominated solution.

[0038] Among all non-dominated solutions, the alternative vehicle combination with the smallest theoretical transportation cost is selected as the first combination; the alternative vehicle combination with the smallest order acceptance cost factor is selected as the second combination.

[0039] As a preferred solution of the adaptive calculation method for waybills based on the dynamic transportation price rule engine described in this application, wherein: the historical waybill data includes the waybill data of each idle vehicle in the transport vehicle combination transporting the same type of goods;

[0040] The theoretical transportation cost is modified to generate a probability distribution of the order price, specifically including:

[0041] Based on historical waybill data and the theoretical unit freight of each idle vehicle, calculate the probability distribution of the theoretical unit freight of each idle vehicle;

[0042] Based on the probability distribution of the theoretical unit freight of each idle vehicle, the theoretical transportation cost of each transport vehicle combination is recalculated to obtain the probability distribution of the order price of each transport vehicle combination.

[0043] As a preferred solution of the adaptive calculation method for shipping orders based on the dynamic transportation price rule engine described in this application, the method for calculating the probability distribution of the theoretical unit freight of any idle vehicle is as follows:

[0044] S10: Extract the unit freight of each waybill of the idle vehicle from the historical waybill data; record the minimum unit freight in the historical waybill data; calculate the standard deviation of the unit freight in the historical waybill data; calculate the rejection rate of the idle vehicle based on the historical waybill data; generate a unit freight range based on the theoretical unit freight of the idle vehicle;

[0045] S20: generating a price threshold for idle vehicles; the generation of the price threshold follows a normal distribution;

[0046] The mean of the normal distribution corresponds to the minimum unit freight in the historical waybill data, and the standard deviation is the standard deviation of the unit freight in the historical waybill data;

[0047] S30: sequentially determining whether each unit freight in the unit freight range has been accepted; specifically, including:

[0048] If the unit freight is less than the price threshold, the unit freight order is rejected; if the unit freight is greater than or equal to the price threshold, the rejection rate is used to simulate whether the unit freight order is accepted;

[0049] S40: Repeat S20 to S30 at least n times, where n is a positive integer; calculate the probability of each unit freight being accepted in the unit freight range to obtain a probability distribution of the theoretical unit freight.

[0050] As a preferred solution of the adaptive calculation method for shipping orders based on the dynamic transportation price rule engine described in this application, generating a recommended price range based on the probability distribution of the order price and the theoretical transportation cost specifically includes:

[0051] Setting an order acceptance probability threshold; determining a first price range based on the probability distribution of the order acceptance prices of the first combination; the first price range includes order acceptance prices in the first combination whose order acceptance probability is greater than the order acceptance probability threshold;

[0052] Determining a second price range based on the probability distribution of the order acceptance prices of the second combination; the second price range includes order acceptance prices of the second combination whose order acceptance probability is greater than the order acceptance probability threshold;

[0053] Merge the first price range and the second price range to obtain a recommended price range;

[0054] The theoretical transportation cost of the first combination is marked as the cost priority price in the recommended price range; the theoretical transportation cost of the second combination is marked as the order acceptance rate priority price in the recommended price range.

[0055] As a preferred solution of the adaptive calculation method for shipping orders based on the dynamic transportation price rule engine described in this application, feedback optimization is performed on the generation of the recommended price range as follows:

[0056] Determine the user's preferred vehicle combination, including:

[0057] Based on the probability distribution of the order prices in the first combination, the probability of accepting the order at the price selected by the user in the first combination is calculated, which is recorded as the first actual probability. Based on the probability distribution of the order prices in the second combination, the probability of accepting the order at the price selected by the user in the second combination is calculated, which is recorded as the second actual probability.

[0058] If only the first actual probability exists or the first actual probability is greater than the second actual probability, the user's preferred vehicle combination is the first combination; if only the second actual probability exists or the first actual probability is less than the second actual probability, the user's preferred vehicle combination is the second combination;

[0059] If the user's preferred vehicle combination is the first combination, the weight coefficient of the theoretical transportation cost when calculating the optimization target value in S200 is increased; otherwise, the weight coefficient of the order acceptance cost factor is increased.

[0060] As a preferred solution of the adaptive calculation method for waybills based on the dynamic transportation price rule engine described in this application, the method of determining the user's preferred vehicle combination further includes:

[0061] If the first actual probability and the second actual probability do not exist, or the first actual probability is equal to the second actual probability, then the absolute value of the difference between the order price selected by the user and the median of the first price range is calculated, and recorded as the first price difference; the absolute value of the difference between the order price selected by the user and the median of the second price range is calculated, and recorded as the second price difference; if the first price difference is greater than the second price difference, the user's preferred vehicle combination is the first combination; otherwise, the user's preferred vehicle combination is the second combination (the absence of a certain actual probability indicates that the price selected by the user is not within the corresponding price range).

[0062] Feedback optimization is performed on the generation of the recommended price range, further comprising: if the user's preferred vehicle combination is the first combination, reducing the lower limit of the unit freight range in S10; otherwise, increasing the upper limit of the unit freight range in S10 and increasing the order acceptance probability threshold.

[0063] In a second aspect, the present application provides a waybill adaptive calculation system based on a dynamic transportation price rule engine, comprising a data acquisition module, a price engine module, a vehicle screening module, a vehicle combination module, a price recommendation module, and a feedback optimization module; wherein:

[0064] The data acquisition module is used to obtain the user's transportation demand data and idle vehicle data;

[0065] The price engine module is equipped with a dynamic transportation price rule engine to calculate the theoretical unit freight rate for each idle vehicle;

[0066] The vehicle screening module is used to screen candidate vehicles based on transportation demand data and idle vehicle data;

[0067] The vehicle combination module outputs the transport vehicle combination through iterative optimization and calculates the theoretical transportation cost of each transport vehicle combination;

[0068] The price recommendation module is used to modify the theoretical transportation cost, generate the probability distribution of the order price for each transportation vehicle combination, and then generate the recommended price range;

[0069] The feedback optimization module performs feedback optimization on the generation of recommended price ranges based on the order price selected by the user.

[0070] Compared with the prior art, the present invention has the following beneficial effects:

[0071] This application uses data on transportation demand and idle vehicles to accurately screen and build vehicle portfolios, increasing the success rate of first-time bid acceptance. By setting price thresholds that follow a normal distribution and conducting multiple rounds of simulated acceptance, and taking into account driver behavior and other environmental factors, the probability distribution of acceptance prices more accurately reflects market conditions, providing users with a more reliable price reference.

[0072] This application uses a genetic algorithm and Pareto optimization to generate transport vehicle combinations, resulting in the combination with the lowest theoretical transport cost and the highest order acceptance rate. This balances transport cost and order acceptance rate, ensuring the feasibility and affordability of recommended prices and meeting the cost and efficiency requirements of different users. Feedback optimization is performed based on the user's selected order price, ensuring that the subsequently generated transport plan is more in line with the user's actual needs and preferences, improving the user experience. BRIEF DESCRIPTION OF THE DRAWINGS

[0073] In order to more clearly illustrate the technical solutions of the embodiments of the present application, the following briefly introduces the drawings required for describing the embodiments. Obviously, the drawings described below are only some embodiments of the present application. For those skilled in the art, other drawings can be derived from these drawings without inventive effort. Among them:

[0074] Figure 1 A flowchart of the adaptive calculation method for shipping orders based on a dynamic transport price rule engine provided by this application;

[0075] Figure 2 Flowchart of the method for generating alternative vehicle combinations provided in this application. DETAILED DESCRIPTION

[0076] The technical solution of the present application is described in detail below through the accompanying drawings and specific embodiments. It should be understood that the embodiments of the present application and the specific features in the embodiments are detailed descriptions of the technical solution of the present application, rather than limitations on the technical solution of the present application. Unless there is a conflict, the embodiments of the present application and the technical features in the embodiments can be combined with each other.

[0077] Example 1

[0078] This embodiment introduces a waybill adaptive calculation method based on a dynamic transportation price rule engine. Figure 1 , the method comprises the following steps:

[0079] Obtain users' transportation demand data and idle vehicle data;

[0080] The transportation demand data includes cargo type, cargo weight, starting point, destination, and deadline. The idle vehicle data includes a list of idle vehicles, each idle vehicle's model, load capacity, current location, idle time, remaining available time, and driver credit score. For any idle vehicle, the remaining available time is the time between the current moment and its next scheduled waybill or its unavailability (e.g., maintenance time). The remaining available time reflects the length of time the vehicle is available to take on new waybill tasks. The idle time is the time elapsed from the completion of the previous waybill to the current moment. The driver credit score is calculated based on the idle vehicle's corresponding waybill completion rate, on-time delivery rate, cargo integrity rate, and violation records. The driver credit score is calculated by obtaining historical waybill data for the idle vehicles. The driver credit score is then assigned based on the data, including the waybill completion rate, on-time delivery rate, cargo integrity rate, and violation records (e.g., temporary price increases) for each idle vehicle. Alternatively, if an existing online waybill platform has a built-in driver credit score scoring function, the driver credit score recorded on the platform can be directly used.

[0081] Loading preset dynamic transportation pricing rules; calculating the theoretical unit freight for each idle vehicle based on the dynamic transportation pricing rules, transportation demand data, and idle vehicle data;

[0082] The theoretical unit freight is the freight for transporting unit weight of goods per unit distance when an idle vehicle executes a waybill issued by a user. The method for calculating the theoretical unit freight based on the dynamic transportation price rule includes:

[0083] Determining a transportation route and a delivery date based on the transportation demand data; wherein the transportation route is determined based on the starting point and the end point; and the delivery date is determined based on the cut-off time;

[0084] Call the weather API to obtain the weather conditions for the transportation route during the delivery date;

[0085] Based on the transportation route, delivery date, weather conditions, cargo type, model and load capacity of the idle vehicle, the theoretical unit freight for each idle vehicle when executing the waybill issued by the user is dynamically calculated.

[0086] The present embodiment calculates the theoretical unit freight by optimizing the dynamic transportation price rules as follows: determining the cargo type, including but not limited to general cargo, perishables, and fragile goods; determining the load capacity and vehicle type, including but not limited to van, flatbed, refrigerated and other vehicles; determining the weather conditions, including but not limited to sunny days, rain, snow, typhoons, etc.; determining the date label of the delivery date, including weekdays, holidays, etc.; determining the characteristic parameters of the transportation route, including the proportion of highways, etc.; through the preset dynamic transportation price rule engine, inputting the cargo type, load capacity and vehicle type, weather conditions, date label of the delivery date, and characteristic parameters of the transportation route, calculates the theoretical unit freight for each idle vehicle when executing the waybill issued by the user.

[0087] Based on the transportation demand data and the idle vehicle data, screening candidate vehicles and building a candidate vehicle set;

[0088] Based on the transportation demand data and idle vehicle data, candidate vehicles are screened, specifically including:

[0089] Set vehicle model requirements based on the cargo type; filter out idle vehicles that meet the vehicle model requirements to form the first candidate set; for example, for raw meat cargo, the vehicle model requirement is a refrigerated vehicle.

[0090] Setting a distance threshold; filtering idle vehicles whose distance between the current location and the starting point is less than the distance threshold to form a second candidate set;

[0091] The maximum transportation time is set based on the deadline; idle vehicles whose remaining available time is greater than the maximum transportation time are screened to form a third alternative set; the preferred method of setting the maximum transportation time in this embodiment is as follows: the basic transportation time is obtained by subtracting the current time from the deadline; a buffer time is set, such as 24 hours, and the basic transportation time is added to the buffer time to obtain the maximum transportation time.

[0092] Setting a credit score threshold; screening idle vehicles whose drivers' credit scores are greater than the credit score threshold to form a fourth candidate set;

[0093] Taking the intersection of the first candidate set, the second candidate set, the third candidate set, and the fourth candidate set to obtain candidate vehicles;

[0094] The candidate vehicle set includes all candidate vehicles and the load capacity, current location, idle time, remaining available time, and theoretical unit freight corresponding to each candidate vehicle.

[0095] Based on the candidate vehicle set and the theoretical unit freight, outputting a transport vehicle combination and a theoretical transport cost of each transport vehicle combination through iterative optimization; specifically including:

[0096] Generate at least m groups of alternative vehicle combinations iteratively through the genetic algorithm, and record the theoretical transportation cost of each alternative vehicle combination; m is a positive integer;

[0097] A non-dominated solution is screened from the alternative vehicle combinations as a transport vehicle combination; the transport vehicle combination includes a first combination and a second combination, wherein the first combination is the transport vehicle combination with the lowest theoretical transport cost, and the second combination is the transport vehicle combination with the highest order acceptance rate.

[0098] Reference Figure 2 , the genetic algorithm specifically includes:

[0099] S100: Generate an idle vehicle combination for receiving orders; assign a task amount to each idle vehicle in each idle vehicle combination to obtain an initial population; the task amount is the cargo weight corresponding to the idle vehicle receiving the order;

[0100] S200: Calculate an optimization target value for each idle vehicle combination; record idle vehicle combinations whose optimization target values ​​are less than a preset optimization target threshold as candidate vehicle combinations; and record the number of candidate vehicle combinations;

[0101] The method for calculating the optimization target value of any idle vehicle combination is as follows:

[0102] The theoretical freight of each idle vehicle is calculated based on the task volume of each idle vehicle and the theoretical unit freight; the theoretical freight sum of each idle vehicle in the idle vehicle combination is calculated to obtain the theoretical transportation cost of the idle vehicle combination; in this embodiment, the total freight of each idle vehicle is obtained as its theoretical freight by multiplying the task volume by the theoretical unit freight and then by the length of the transportation route;

[0103] The idle distance of each idle vehicle is calculated based on the current position of each idle vehicle; the preferred method for calculating the idle distance in this implementation is to call the electronic map API, input the current position of the idle vehicle as its starting point, input the starting point in the transportation demand data as the end point of the idle vehicle, and generate the shortest path as the idle distance of the idle vehicle.

[0104] The order acceptance cost factor of each idle vehicle is calculated based on the idle distance, idle time, and remaining available time of each idle vehicle; the sum of the order acceptance cost factors of each idle vehicle in the idle vehicle combination is calculated to obtain the order acceptance cost factor of the idle vehicle combination; the preferred method of calculating the order acceptance cost factor of any idle vehicle in this embodiment is as follows: the idle distance, idle time, and remaining available time are normalized and then weighted and summed to obtain the order acceptance cost factor, wherein the weight of the idle distance is , the weight of idle time is -1 times The weight of the remaining available time is -1 times ; 、 、 The sum of these factors is 1 and all factors are greater than 0. The greater the idle distance, the higher the driver's cost of accepting orders and the larger the acceptance cost factor. The longer the idle time, the greater the driver's willingness to accept orders and the smaller the acceptance cost factor. The longer the remaining available time, the less pressure the driver faces to accept orders and the smaller the acceptance cost factor. The lower the acceptance cost factor, the greater the driver's willingness to accept orders and the higher the acceptance rate.

[0105] The theoretical transportation cost of the idle vehicle combination is normalized and then weighted together with the order acceptance cost factor of the idle vehicle combination to obtain the optimization target value of the idle vehicle combination. In this embodiment, when performing the initial shipping order calculation for each user and calculating the optimization target value, the weight of the order acceptance cost factor and the sum of the theoretical transportation cost is preferably 0.5.

[0106] This embodiment uses an iterative genetic algorithm to minimize the optimization objective, ensuring that the candidate vehicle combinations balance transportation costs and order acceptance rates to a certain extent. This avoids overly extreme combinations, such as those that overly compress costs or overly prioritize order acceptance rates while completely ignoring transportation costs, and ensures the feasibility of idle vehicle combinations. By adjusting the weighting of the order acceptance cost factor and the theoretical transportation cost, it is possible to explore different idle vehicle combinations.

[0107] S300: Generate the next generation population through selection operation, crossover operation and mutation operation;

[0108] S400: iteratively executing S200-S300 until the number of candidate vehicle combinations is greater than or equal to m, and then stopping the iteration;

[0109] S500: Outputting alternative vehicle combinations and theoretical transportation costs of each alternative vehicle combination.

[0110] In the genetic algorithm, the common constraints of each generation of population include: the sum of the loads of all idle vehicles in any idle vehicle combination is not less than the cargo weight. times, and not more than the weight of the goods times; among them, is the preset lower load limit multiple, is a multiple of the preset upper load limit; in this embodiment, the preferred is 1.05, It is 1.2, ensuring that any idle vehicle combination can complete the orders placed by users without overloading, while retaining a certain load margin to avoid excessive transportation pressure.

[0111] This embodiment uses Pareto optimization to select non-dominated solutions from alternative vehicle combinations as the transport vehicle combination, specifically including:

[0112] Each alternative vehicle combination is considered a candidate solution. All candidate solutions form a solution space. The theoretical transportation cost and order acceptance cost factor corresponding to each candidate solution in the solution space are recorded. Non-dominated solutions in the solution space are iteratively screened, specifically including: if for any candidate solution A, there exists at least one candidate solution B such that the theoretical transportation cost and order acceptance cost factor of candidate solution B are both no greater than those of candidate solution A, and at least one of the theoretical transportation cost and order acceptance cost factors of candidate solution B is less than the corresponding item of candidate solution A, then candidate solution B dominates candidate solution A; otherwise, candidate solution A is a non-dominated solution.

[0113] Among all non-dominated solutions, the alternative vehicle combination with the smallest theoretical transportation cost is selected as the first combination; the alternative vehicle combination with the smallest order acceptance cost factor is selected as the second combination.

[0114] Theoretical transportation cost and order acceptance cost factors are two optimization objectives with a significant trade-off, making them suitable for Pareto optimization. For example, selecting a vehicle with a low theoretical unit freight rate, coupled with long idle distances (necessitating dispatch from a distance), will result in a lower order acceptance rate. Meanwhile, drivers with long idle times are more willing to accept orders with slightly lower freight rates to alleviate idle pressure.

[0115] Obtain historical waybill data; modify the theoretical transportation cost based on the historical waybill data to generate a probability distribution of the order price;

[0116] The historical waybill data includes waybill data of each idle vehicle in the transport vehicle combination transporting the same type of goods;

[0117] The theoretical transportation cost is modified to generate a probability distribution of the order price, specifically including:

[0118] Based on historical waybill data and the theoretical unit freight of each idle vehicle, the probability distribution of the theoretical unit freight of each idle vehicle is calculated.

[0119] Based on the probability distribution of the theoretical unit freight of each idle vehicle, the theoretical transportation cost of each transport vehicle combination is recalculated to obtain the probability distribution of the order price of each transport vehicle combination.

[0120] The probability distribution of the theoretical unit freight for any idle vehicle is calculated as follows:

[0121] S10: Extract the unit freight of each waybill for idle vehicles in the historical waybill data; record the minimum unit freight in the historical waybill data; calculate the standard deviation of the unit freight in the historical waybill data; calculate the rejection rate of idle vehicles based on the historical waybill data; generate a unit freight range based on the theoretical unit freight of idle vehicles; in this embodiment, the lower limit of the unit freight range is preferably 0.8 times the theoretical unit freight, and the upper limit is 1.1 times the theoretical unit freight.

[0122] S20: generating a price threshold for idle vehicles; the generation of the price threshold follows a normal distribution;

[0123] The mean of the normal distribution corresponds to the minimum unit freight in the historical waybill data, and the standard deviation is the standard deviation of the unit freight in the historical waybill data;

[0124] S30: sequentially determining whether each unit freight in the unit freight range has been accepted; specifically, including:

[0125] If the unit freight rate is less than the price threshold, the order is rejected. If the unit freight rate is greater than or equal to the price threshold, the rejection rate is used to simulate whether the unit freight order will be accepted. For example, if the rejection rate is 20%, the probability of the unit freight order being accepted is set to 80% during the simulation.

[0126] S40: Repeat S20 to S30 at least n times, where n is a positive integer; calculate the probability of each unit freight being accepted in the unit freight range to obtain a probability distribution of the theoretical unit freight.

[0127] By setting a price threshold that follows a normal distribution and conducting multiple rounds of simulated order acceptance, noise interference including driver behavior and other environmental factors is injected into the simulation of order acceptance probability, so that the probability distribution of order acceptance price can reflect the actual situation.

[0128] In this embodiment, the total freight of each idle vehicle is obtained as its theoretical freight by multiplying the task volume by the theoretical unit freight and then by the length of the transportation route. Similarly, different theoretical freights of the idle vehicle are obtained by multiplying different unit freights by the task volume and then by the length of the transportation route. The probability distribution of the theoretical freight is the same as the probability distribution of the theoretical unit freight. By using each theoretical freight and the corresponding probability of each idle vehicle, the theoretical freight of all idle vehicles in the transportation vehicle combination is summed to obtain the probability of the sum of each theoretical freight, that is, the probability of accepting an order at each order price.

[0129] Generate a recommended price range based on the probability distribution of the order price and the theoretical transportation cost; specifically including:

[0130] Setting an order acceptance probability threshold; determining a first price range based on the probability distribution of the order acceptance prices of the first combination; the first price range includes order acceptance prices in the first combination whose order acceptance probability is greater than the order acceptance probability threshold;

[0131] Determining a second price range based on the probability distribution of the order acceptance prices of the second combination; the second price range includes order acceptance prices of the second combination whose order acceptance probability is greater than the order acceptance probability threshold;

[0132] Merge the first price range and the second price range to obtain a recommended price range;

[0133] The theoretical transportation cost of the first combination is marked as the cost priority price in the recommended price range; the theoretical transportation cost of the second combination is marked as the order acceptance rate priority price in the recommended price range.

[0134] Based on the order price selected by the user, feedback optimization is performed on the generation of the recommended price range.

[0135] The order price selected by the user is the actual order price issued by the user after referring to the recommended price range. Feedback optimization is performed on the generation of the recommended price range as follows:

[0136] Determine the user's preferred vehicle combination, including:

[0137] Based on the probability distribution of the order prices in the first combination, the probability of accepting the order at the price selected by the user in the first combination is calculated, which is recorded as the first actual probability. Based on the probability distribution of the order prices in the second combination, the probability of accepting the order at the price selected by the user in the second combination is calculated, which is recorded as the second actual probability.

[0138] If only the first actual probability exists or the first actual probability is greater than the second actual probability, the user's preferred vehicle combination is the first combination; if only the second actual probability exists or the first actual probability is less than the second actual probability, the user's preferred vehicle combination is the second combination;

[0139] If neither the first actual probability nor the second actual probability exists, or if the first actual probability is equal to the second actual probability, then the absolute value of the difference between the order price selected by the user and the median of the first price range is calculated, recorded as the first price difference; the absolute value of the difference between the order price selected by the user and the median of the second price range is calculated, recorded as the second price difference. If the first price difference is greater than the second price difference, the user's preferred vehicle combination is the first combination; otherwise, the user's preferred vehicle combination is the second combination (the absence of a particular actual probability indicates that the price selected by the user is not within the corresponding price range).

[0140] If the user's preferred vehicle combination is the first combination, the weight coefficient of the theoretical transportation cost is increased when calculating the optimization target value in S200. Otherwise, the weight coefficient of the order acceptance cost factor is increased. For example, when the user's preferred vehicle combination is the first combination the next time a shipping price is calculated, indicating that the user prioritizes low costs, the weight coefficient of the theoretical transportation cost is increased from 0.5 to 0.6 when calculating the optimization target value during the genetic algorithm iterations. Accordingly, the weight coefficient of the order acceptance cost factor is decreased from 0.5 to 0.4, thereby strengthening the genetic algorithm's search for alternative vehicle combinations with lower theoretical transportation costs. If the user's preferred vehicle combination is the first combination, the lower limit of the unit freight range is decreased to guide the genetic algorithm to explore recommended prices with lower costs. If the user's preferred vehicle combination is the second combination, the upper limit of the unit freight range is increased, and the order acceptance probability threshold is increased to guide the genetic algorithm to explore recommended prices with higher order acceptance rates.

[0141] Example 2

[0142] This embodiment is the second embodiment of the present application. Based on the same inventive concept as the first embodiment, this embodiment introduces a waybill adaptive calculation system based on a dynamic transportation price rule engine, including a data acquisition module, a price engine module, a vehicle screening module, a vehicle combination module, a price recommendation module, and a feedback optimization module. Specifically:

[0143] The data acquisition module is used to obtain users' transportation demand data and idle vehicle data. Transportation demand data includes cargo type, weight, origin, destination, and deadline. Idle vehicle data includes a list of idle vehicles, along with each vehicle's model, load capacity, current location, idle time, remaining available time, and driver credit score, providing basic data support for subsequent calculations and screening.

[0144] The price engine module is equipped with a dynamic transportation price rule engine, which is used to calculate the theoretical unit freight of each idle vehicle; the dynamic transportation price rule built into the dynamic transportation price rule engine calculates the theoretical unit freight of each idle vehicle based on the transportation route, delivery date, weather conditions, cargo type, idle vehicle model and load capacity.

[0145] The vehicle screening module is used to screen alternative vehicles based on transportation demand data and idle vehicle data; by setting vehicle model requirements, distance thresholds, maximum transportation time and credit score thresholds, idle vehicles that meet the conditions are screened out and an alternative vehicle set is constructed, which contains all alternative vehicles, their related attributes and theoretical unit freight.

[0146] The vehicle combination module outputs the transport vehicle combination through iterative optimization and calculates the theoretical transportation cost of each transport vehicle combination; this module uses a genetic algorithm to iteratively generate at least m groups of alternative vehicle combinations, calculates the theoretical transportation cost and order acceptance cost factor of each combination, and then obtains the optimization target value; uses Pareto optimization to screen out non-dominated solutions from the alternative vehicle combinations, and determines the first combination with the lowest theoretical transportation cost and the second combination with the highest order acceptance rate.

[0147] The price recommendation module modifies theoretical transportation costs, generates a probability distribution of order prices for each vehicle combination, and then generates a recommended price range. This module calculates the probability distribution of theoretical unit freight for each idle vehicle, and then recalculates the theoretical transportation cost for each vehicle combination to obtain a probability distribution of order prices. A threshold for order acceptance probability is set, and the first and second price ranges are determined and merged to form a recommended price range. The recommended price range is then labeled with the cost-priority and acceptance rate-priority pricing.

[0148] The feedback optimization module provides feedback optimization for generating recommended price ranges based on the user's selected order price. This module calculates the difference between the user's selected order price and the median of the first and second price ranges to determine the user's preferred vehicle combination. Based on this preferred vehicle combination, it adjusts the weighting coefficients of the theoretical transportation cost and order acceptance cost factors used in the genetic algorithm in the vehicle combination module to optimize subsequent shipping order calculation results and ensure that subsequent recommended prices are more in line with the user's psychological expectations.

[0149] The specific functional implementation of each of the above modules can be found in the relevant content of the adaptive calculation method for waybills based on a dynamic transportation price rule engine described in Example 1, and will not be elaborated on here.

[0150] Those skilled in the art will appreciate that the embodiments of the present application may be provided as methods, systems, or computer program products. Therefore, the present application may take the form of an entirely hardware embodiment, an entirely software embodiment, or an embodiment combining software and hardware. Furthermore, the present application may take the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to magnetic disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0151] The above describes the embodiments of the present application in conjunction with the accompanying drawings, but the present application is not limited to the above-mentioned specific implementation methods. The above-mentioned specific implementation methods are merely illustrative and not restrictive. Under the guidance of this application, ordinary technicians in this field can also make many forms without departing from the purpose and scope of protection of this application, all of which are protected by this application.

Claims

1. The adaptive calculation method for shipping orders based on a dynamic transportation price rule engine is characterized by: The following steps are involved: Obtain users' transportation demand data and idle vehicle data; Loading preset dynamic transportation pricing rules; calculating the theoretical unit freight for each idle vehicle based on the dynamic transportation pricing rules, transportation demand data, and idle vehicle data; Based on the transportation demand data and the idle vehicle data, screening candidate vehicles and building a candidate vehicle set; Based on the candidate vehicle set and the theoretical unit freight, output the transport vehicle combination and the theoretical transport cost of each transport vehicle combination through iterative optimization, specifically including: Generate at least m groups of alternative vehicle combinations iteratively through the genetic algorithm, and record the theoretical transportation cost of each alternative vehicle combination; m is a positive integer; Selecting a non-dominated solution from the candidate vehicle combinations as a transport vehicle combination; specifically comprising: Each alternative vehicle combination is considered a candidate solution. All candidate solutions form a solution space. The theoretical transportation cost and order acceptance cost factor corresponding to each candidate solution in the solution space are recorded. Non-dominated solutions in the solution space are iteratively screened, specifically including: if for any candidate solution A, there exists at least one candidate solution B such that the theoretical transportation cost and order acceptance cost factor of candidate solution B are both no greater than those of candidate solution A, and at least one of the theoretical transportation cost and order acceptance cost factors of candidate solution B is less than the corresponding item of candidate solution A, then candidate solution B dominates candidate solution A; otherwise, candidate solution A is a non-dominated solution. The transport vehicle combination includes a first combination and a second combination; among all non-dominated solutions, the alternative vehicle combination with the minimum theoretical transport cost is selected as the first combination; and the alternative vehicle combination with the minimum order acceptance cost factor is selected as the second combination; Obtain historical waybill data; modify the theoretical transportation cost based on the historical waybill data to generate a probability distribution of the order price; generating a recommended price range based on the probability distribution of the order price and the theoretical transportation cost; Based on the order price selected by the user, feedback optimization is performed on the generation of the recommended price range.

2. The adaptive calculation method for shipping orders based on a dynamic transportation pricing rule engine according to claim 1, characterized in that: The transportation demand data includes cargo type, cargo weight, starting point, destination, and deadline; the idle vehicle data includes a list of idle vehicles, the model of each idle vehicle, load capacity, current location, idle time, remaining available time, and driver credit score; The theoretical unit freight is the freight for transporting unit weight of goods per unit distance when an idle vehicle executes a waybill issued by a user; Methods for calculating theoretical unit freight based on dynamic transportation price rules include: Determining a transportation route and a delivery date based on the transportation demand data; wherein the transportation route is determined based on the starting point and the end point; and the delivery date is determined based on the deadline; Call the weather API to obtain the weather conditions for the transportation route during the delivery date; Based on the transportation route, delivery date, weather conditions, cargo type, model and load capacity of the idle vehicle, the theoretical unit freight for each idle vehicle when executing the waybill issued by the user is dynamically calculated.

3. The adaptive calculation method for shipping orders based on a dynamic transportation price rule engine according to claim 2, characterized in that: The genetic algorithm specifically includes: S100: Generate an idle vehicle combination for receiving orders; assign a task amount to each idle vehicle in each idle vehicle combination to obtain an initial population; the task amount is the cargo weight corresponding to the idle vehicle receiving the order; S200: Calculate an optimization target value for each idle vehicle combination; record idle vehicle combinations whose optimization target values ​​are less than a preset optimization target threshold as candidate vehicle combinations; and record the number of candidate vehicle combinations; S300: Generate the next generation population through selection operation, crossover operation and mutation operation; S400: iteratively executing S200-S300 until the number of candidate vehicle combinations is greater than or equal to m, and then stopping the iteration; S500: Outputting the alternative vehicle combinations and the theoretical transportation cost of each alternative vehicle combination; In the genetic algorithm, the common constraints of each generation of population include: the sum of the loads of all idle vehicles in any idle vehicle combination is not less than the cargo weight. times, and not more than the weight of the goods times; among them, is the preset lower load limit multiple, It is a multiple of the preset upper load limit.

4. The adaptive calculation method for shipping orders based on a dynamic transportation price rule engine according to claim 3, characterized in that: The method for calculating the optimization target value of any idle vehicle combination is as follows: Calculating the theoretical freight of each idle vehicle based on the task volume of each idle vehicle and the theoretical unit freight; calculating the sum of the theoretical freight of each idle vehicle in the idle vehicle combination to obtain the theoretical transportation cost of the idle vehicle combination; Calculating the idle distance of each idle vehicle based on the current position of each idle vehicle; Calculate the order acceptance cost factor for each idle vehicle based on the idle distance, idle time, and remaining available time of each idle vehicle; Calculate the sum of the order acceptance cost factors of each idle vehicle in the idle vehicle combination to obtain the order acceptance cost factor of the idle vehicle combination; The theoretical transportation cost of the idle vehicle combination is normalized and then weighted summed with the order acceptance cost factor of the idle vehicle combination to obtain the optimization target value of the idle vehicle combination.

5. The adaptive calculation method for shipping orders based on a dynamic transportation price rule engine according to claim 4, characterized in that: The historical waybill data includes waybill data of each idle vehicle in the transport vehicle combination transporting the same type of goods; The theoretical transportation cost is modified to generate a probability distribution of the order price, specifically including: Based on historical waybill data and the theoretical unit freight of each idle vehicle, calculate the probability distribution of the theoretical unit freight of each idle vehicle; Based on the probability distribution of the theoretical unit freight of each idle vehicle, the theoretical transportation cost of each transport vehicle combination is recalculated to obtain the probability distribution of the order price of each transport vehicle combination.

6. The adaptive calculation method for shipping orders based on a dynamic transportation price rule engine according to claim 5, characterized in that: The probability distribution of the theoretical unit freight for any idle vehicle is calculated as follows: S10: Extract the unit freight of each waybill in the historical waybill data of the idle vehicle; record the minimum value of the unit freight in the historical waybill data; Calculate the standard deviation of unit freight in historical waybill data; Calculate the rejection rate of idle vehicles based on historical shipping order data; generate unit freight rate ranges based on the theoretical unit freight rates of idle vehicles; S20: generating a price threshold for idle vehicles; the generation of the price threshold follows a normal distribution; The mean of the normal distribution corresponds to the minimum unit freight in the historical waybill data, and the standard deviation is the standard deviation of the unit freight in the historical waybill data; S30: sequentially determining whether each unit freight in the unit freight range has been accepted; specifically, including: If the unit freight is less than the price threshold, the unit freight order is rejected; if the unit freight is greater than or equal to the price threshold, the rejection rate is used to simulate whether the unit freight order is accepted; S40: Repeat S20 to S30 at least n times, where n is a positive integer; calculate the probability of each unit freight being accepted in the unit freight range to obtain a probability distribution of the theoretical unit freight.

7. The adaptive calculation method for shipping orders based on a dynamic transportation price rule engine according to claim 6, characterized in that: Generate a recommended price range based on the probability distribution of the order price and the theoretical transportation cost, specifically including: Setting an order acceptance probability threshold; determining a first price range based on the probability distribution of the order acceptance prices of the first combination; the first price range includes order acceptance prices in the first combination whose order acceptance probability is greater than the order acceptance probability threshold; Determining a second price range based on the probability distribution of the order acceptance prices of the second combination; the second price range includes order acceptance prices of the second combination whose order acceptance probability is greater than the order acceptance probability threshold; Merge the first price range and the second price range to obtain a recommended price range; The theoretical transportation cost of the first combination is marked as the cost priority price in the recommended price range; the theoretical transportation cost of the second combination is marked as the order acceptance rate priority price in the recommended price range.

8. The adaptive calculation method for shipping orders based on a dynamic transportation price rule engine according to claim 7, characterized in that: Feedback optimization is performed on the generation of the recommended price range as follows: Determine the user's preferred vehicle combination, including: Based on the probability distribution of the order prices in the first combination, the probability of accepting the order at the price selected by the user in the first combination is calculated, which is recorded as the first actual probability. Based on the probability distribution of the order prices in the second combination, the probability of accepting the order at the price selected by the user in the second combination is calculated, which is recorded as the second actual probability. If only the first actual probability exists or the first actual probability is greater than the second actual probability, the user's preferred vehicle combination is the first combination; if only the second actual probability exists or the first actual probability is less than the second actual probability, the user's preferred vehicle combination is the second combination; If the user's preferred vehicle combination is the first combination, the weight coefficient of the theoretical transportation cost when calculating the optimization target value in S200 is increased; otherwise, the weight coefficient of the order acceptance cost factor is increased.

9. The adaptive calculation method for shipping orders based on a dynamic transportation price rule engine according to claim 8, characterized in that: Determining the user's preferred vehicle combination further includes: If the first actual probability and the second actual probability do not exist, or the first actual probability is equal to the second actual probability, then the absolute value of the difference between the order price selected by the user and the median of the first price range is calculated, recorded as the first price difference; the absolute value of the difference between the order price selected by the user and the median of the second price range is calculated, recorded as the second price difference; if the first price difference is greater than the second price difference, the user's preferred vehicle combination is the first combination; otherwise, the user's preferred vehicle combination is the second combination; Feedback optimization is performed on the generation of the recommended price range, further comprising: if the user's preferred vehicle combination is the first combination, reducing the lower limit of the unit freight range in S10; otherwise, increasing the upper limit of the unit freight range in S10 and increasing the order acceptance probability threshold.

10. A waybill adaptive calculation system, for implementing the waybill adaptive calculation method according to any one of claims 1 to 9, characterized in that: It includes data acquisition module, price engine module, vehicle screening module, vehicle combination module, price recommendation module, and feedback optimization module; among them: The data acquisition module is used to obtain the user's transportation demand data and idle vehicle data; The price engine module is equipped with a dynamic transportation price rule engine to calculate the theoretical unit freight rate for each idle vehicle; The vehicle screening module is used to screen candidate vehicles based on transportation demand data and idle vehicle data; The vehicle combination module outputs the transport vehicle combination through iterative optimization and calculates the theoretical transportation cost of each transport vehicle combination; The price recommendation module is used to modify the theoretical transportation cost, generate the probability distribution of the order price for each transportation vehicle combination, and then generate the recommended price range; The feedback optimization module performs feedback optimization on the generation of recommended price ranges based on the order price selected by the user.

Images