Order management method based on cloud computing service

Abstract

The present invention relates to the technical field of order management. Disclosed is an order management method based on a cloud computing service. The method comprises: a cloud platform receiving a passenger location sent by a passenger terminal; acquiring real-time traffic data, defining a screening radius, and generating an order dispatching area for the passenger location; receiving driver locations sent by all driver terminals within the order dispatching area; acquiring road traffic information, fusing the road traffic information, the passenger location, and all the driver locations, calculating a shortest pick-up time from each driver location to the passenger location, and generating a time ranking list; and sending an order to the driver terminal with the shortest pick-up time, and if the driver terminal does not accept the order, sequentially dispatching the order to the driver terminals on the basis of the time ranking list.

Description

A cloud computing-based order management method Technical Field

[0001] This invention relates to the field of order management technology, and more specifically, to an order management method based on cloud computing services. Background Technology

[0002] With the rise of the sharing economy, ride-hailing services have become an important mode of urban transportation. In the operation of ride-hailing platforms, real-time order scheduling is the key to improving service efficiency. Traditional scheduling methods usually rely on fixed area divisions or simple distance matching algorithms, which cannot fully consider dynamic factors such as real-time traffic conditions and driver trajectories. This results in problems such as uneven order distribution, unreasonable scheduling, and long waiting times for users, which affect passenger experience and driver income. Therefore, there is an urgent need for an intelligent ride-hailing order management method.

[0003] Patent CN117522530B discloses a method for managing ride-hailing orders. The method includes: when a ride-hailing management center receives a user's ride order via a wireless network, it first locates the user's position, then generates a dispatch area A for the user with the user's location as the center and a preset radius r1. It then obtains the number of ride-hailing vehicles in dispatch area A that are accepting orders. If the number of vehicles in dispatch area A that are accepting orders is greater than 1, it obtains multiple data points for all ride-hailing vehicles, sorts all vehicles based on a sorting algorithm, and assigns the ride order to the vehicle ranked first. This effectively improves the uniformity of order dispatching by the ride-hailing management center and the efficiency of ride order management.

[0004] However, while the aforementioned technologies can manage ride-hailing orders, the dispatching process primarily considers various data points, including the number of orders accepted within a specified time period, mileage index, vehicle idle time, and vehicle breakdown frequency. It does not take into account the time it takes for a ride-hailing vehicle to reach the passenger's location. This results in drivers having to travel longer distances after accepting an order to reach the passenger, increasing empty-running rates and reducing operational efficiency. Simultaneously, it prolongs passenger waiting times, impacting the travel experience. Especially during peak hours or in inclement weather, prolonged waiting times can lead to user churn and reduce platform user stickiness.

[0005] In view of this, the present invention proposes an order management method based on cloud computing services to solve the above problems. Summary of the Invention

[0006] To overcome the aforementioned deficiencies of the prior art and achieve the above objectives, the present invention provides the following technical solution: an order management method based on cloud computing services, applied in a cloud platform, comprising:

[0007] S1: The cloud platform receives the passenger's location sent by the passenger terminal;

[0008] S2: The cloud platform acquires real-time traffic data, defines a filtering radius using a dynamic radius algorithm based on the real-time traffic data, and generates a dispatch area based on the passenger's location based on the filtering radius;

[0009] S3: The cloud platform receives the driver locations sent by all drivers within the dispatch area;

[0010] S4: The cloud platform acquires road traffic information, integrates road traffic information, passenger location and all driver locations, calculates the shortest pick-up time from each driver location to the passenger location, and generates a time sorting table;

[0011] S5: The cloud platform sends the order to the driver with the shortest pick-up time. If the driver does not accept the order, the cloud platform will assign orders to the drivers in order according to the time sorting table.

[0012] Furthermore, the passenger location is the geographical coordinate of the passenger, and the driver location is the geographical coordinate of the driver;

[0013] Real-time traffic data includes traffic parameters for each road segment within the filtering range, including traffic density and traffic speed; the filtering range is a circular area with the passenger's location as the center and a preset filtering distance as the radius;

[0014] The steps to define the filter radius include:

[0015] Step S201: Define the iteration threshold a, step size factor b, adjustment factor d, and intensity factor g;

[0016] Step S202: Preset radius range;

[0017] Step S203: Construct a population S, which includes m individuals. The position of each individual corresponds one-to-one with the value within the radius range. The initial iteration number t of the population S is 0.

[0018] Step S204: Define the improvement function;

[0019] Step S205: Divide the population S into h subpopulations, each subpopulation containing k individuals, where m = hk;

[0020] Step S206: Calculate the flight probability for each individual and determine the corresponding flight mode;

[0021] Step S207: Update the location of each individual;

[0022] Step S208: Calculate the attraction coefficient for each individual and update the position of each individual again;

[0023] Step S209: Compare the iteration number t with the iteration threshold a; if t≥a, proceed to step S210; if t<a, let t=t+1 and return to step S206.

[0024] Step S210: Calculate the degree of improvement for each individual, and use the value of the individual with the highest degree of improvement as the screening radius.

[0025] Further, in step S203, the population S = {X1, X2, ..., X} m}, X m Let m be the m-th individual; the position of each individual in the population S is defined in a one-dimensional search space, the range of which is the radius; the expression for the position of each individual in the population S is: In the formula, Let P be the position of the i-th individual. i Let P be the random coefficient of the i-th individual. i ∈[0,1], i∈[1,m];

[0026] In step S204, the expression for the improvement degree function is: In the formula, f represents the degree of improvement, jd represents the order acceptance time, and ds represents the waiting time. The order acceptance time is the time elapsed from when a passenger places an order to when the driver accepts the order, and the waiting time is the time elapsed from when a passenger places an order to when the driver arrives at the passenger's location. The order acceptance time and the waiting time are obtained by using real-time traffic data and the numerical values ​​corresponding to individual locations as analysis data. The analysis data is input into a trained first-time model to predict the corresponding order acceptance time, and the analysis data is input into a trained second-time model to predict the corresponding waiting time. Both the first-time model and the second-time model are deep neural network models.

[0027] Furthermore, in step S205, the method for dividing the population S into h subpopulations includes:

[0028] Calculate the improvement degree for each individual in population S and sort them from largest to smallest. Assign an incremental number to each individual according to the ascending order, with the sequence number ranging from [1, m]. For each of the h subpopulations, perform a modulo operation on the number of each individual to obtain the corresponding sub-number. The expression for the sub-number is: l = u % h; where l is the sub-number, u is the number, and % is the modulo function. If the sub-number is not 0, the corresponding individual is assigned to the l-th subpopulation; if the sub-number is 0, the corresponding individual is assigned to the h-th subpopulation.

[0029] In step S206, the expression for the flight probability is: In the formula, Let i be the flight probability corresponding to the i-th individual. f represents the degree of improvement corresponding to the optimal individual. i t Let i be the degree of improvement corresponding to the i-th individual. The worst individual is the one with the greatest improvement in population S; the best individual is the one with the greatest improvement in population S, and the worst individual is the one with the least improvement in population S.

[0030] Methods for determining an individual's corresponding flight mode include:

[0031] A preset probability threshold is set, and the flight probability of each individual is compared with the probability threshold. If the flight probability is greater than or equal to the probability threshold, the flight mode of the corresponding individual is global flight; if the flight probability is less than the probability threshold, the flight mode of the corresponding individual is local flight.

[0032] Furthermore, in step S207, the method for updating the position of each individual includes:

[0033] If the flight mode corresponding to an individual is global flight, the methods for updating the location include:

[0034] In the formula, This represents the position of the i-th individual after the update. b represents the position of the i-th individual before the update. t Let be the step size factor in the t-th iteration. For the optimal position of the individual, d t Let be the adjustment factor in the t-th iteration. C1 is a random number between [0,1]. Let g be the flight speed of the i-th individual. t Let be the intensity factor in the t-th iteration. Let be the flight speed of the i-th individual in the previous iteration;

[0035] The expression for flight speed is: In the formula, C2 is a random number between [0,1].

[0036] If the flight mode corresponding to an individual is local flight, the methods for updating the location include:

[0037] In the formula, Let represent the position of the optimal individual in the subpopulation containing the i-th individual. The optimal individual is the one with the greatest improvement in the subpopulation.

[0038] Further, in step S208, the expression for the attraction coefficient is: In the formula, Let be the attraction coefficient between the i'th individual and the i-th individual in the subpopulation containing the i-th individual, and exp be an exponential function. Let be the degree of improvement corresponding to the i′ individual in the subpopulation of the i-th individual, i≠i′, i′∈[1,k];

[0039] The methods for updating the location of each individual again include:

[0040] In the formula, This is the position of the i-th individual after the update. To update the position of the i′th individual in the subpopulation of the i-th individual before the update;

[0041] The dispatch area is a circular area with the passenger's location as the center and the filtering radius as the radius.

[0042] Furthermore, the method for calculating the shortest pick-up time from each driver's location to the passenger's location includes:

[0043] Get all intersections within the dispatch area; based on the passenger location, driver location, and intersection, obtain P pick-up routes from each driver location to the passenger location; calculate the pick-up time corresponding to each pick-up route, compare the pick-up times corresponding to each driver location, and take the shortest pick-up time as the shortest pick-up time for the corresponding driver location.

[0044] Furthermore, the step of obtaining P pick-up routes from each driver's location to the passenger's location includes:

[0045] Step S401: Randomly select a driver's position that is not marked as a selected position and mark it as the current position;

[0046] Step S402: Explore all adjacent intersections at the current location, randomly select one of the adjacent intersections as the successor intersection, and mark it as the selected intersection;

[0047] Step S403: Determine whether the passenger's location is adjacent to the next intersection. If so, use the current location, the selected intersection, and the passenger's location as a pick-up route. If not, proceed to step S404.

[0048] Step S404: Explore all adjacent intersections of the successor intersection, randomly select one of the adjacent intersections that is not marked as a selected intersection as the successor intersection, and mark it as a selected intersection;

[0049] Step S405: Determine whether the passenger's location is adjacent to a subsequent intersection. If so, use the current location, the selected intersection, and the passenger's location as a pick-up route. If not, proceed to step S406.

[0050] Step S406: Determine whether all adjacent intersections of the subsequent intersection have been marked as selected intersections. If yes, demark all intersections marked as selected intersections in step S404 and proceed to step S407. If not, return to step S404.

[0051] Step S407: Repeat steps S402 to S406 until P pick-up routes are obtained, then the loop ends and proceed to step S408;

[0052] Step S408: Repeat steps S401 to S407 until all driver positions are marked as the current position. The loop ends, and P pick-up routes from each driver position to the passenger position are obtained.

[0053] Furthermore, the method for calculating the pick-up time corresponding to each pick-up route includes:

[0054] Each pick-up route is represented by a set of intersections, with each set corresponding to a specific pick-up route. Two adjacent intersections within each set are grouped into an adjacent set, and the corresponding road segments are identified and marked as pick-up segments. Traffic parameters for each pick-up segment are retrieved from real-time traffic data and marked as pick-up parameters. Road traffic information, including path length and traffic light status data, is acquired. Path length data includes the path length for each pick-up route. Traffic light status data includes the light status and duration at each intersection along each pick-up route. Different numerical labels are assigned to different light statuses and marked as light labels. The light status data is then replaced with the corresponding light labels. The pick-up parameters, path length, and traffic light status data for each pick-up route are treated as a set of computational data, with each set corresponding to a specific pick-up route. Each set of computational data is input into a trained time prediction model, which is a deep neural network model, to predict the corresponding pick-up time.

[0055] Methods for generating time-sorted tables include:

[0056] Sort all the shortest pick-up times from shortest to longest to generate a time sorting table.

[0057] Furthermore, if no driver accepts the order, orders are assigned to drivers in ascending order according to the time sorting table. If no driver accepts the order, the filtering radius is multiplied by a preset expansion coefficient to obtain the expansion radius. Based on the expansion radius, an order assignment area is generated again for the passenger's location, and orders are reassigned.

[0058] The technical effects and advantages of the order management method based on cloud computing services of this invention are as follows:

[0059] By acquiring passenger location and real-time traffic data, a dynamic radius algorithm is used to adaptively generate dispatch areas, enabling dynamic adjustment of the dispatch range and improving dispatch accuracy. The shortest pick-up time from each driver's location to the passenger's location is calculated in real time, prioritizing dispatches to drivers closest to the passenger, thereby reducing passenger waiting time and driver empty-running rate, and effectively improving user experience. By fully utilizing cloud computing and big data analytics, intelligent optimization and dynamic adjustment of order dispatch are achieved, improving overall operational efficiency and adapting to travel demands in the context of the sharing economy. Attached Figure Description

[0060] Figure 1 is a flowchart of an order management method based on cloud computing services according to Embodiment 1 of the present invention; Detailed Implementation

[0061] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0062] Example 1

[0063] Please refer to Figure 1. This embodiment describes an order management method based on cloud computing services. The method includes:

[0064] S1: The cloud platform receives the passenger's location sent by the passenger terminal.

[0065] A cloud platform is a service provider based on cloud computing technology, which typically includes components such as servers, storage, databases, and applications to process and manage data; the passenger end is the application used by passengers.

[0066] The passenger location is the geographical coordinates of the passenger's location. The passenger location is obtained by the passenger requesting access to the GPS of the first device, which is the mobile device used by the passenger.

[0067] S2: The cloud platform acquires real-time traffic data, defines a filtering radius using a dynamic radius algorithm based on the real-time traffic data, and generates a dispatch area based on the passenger's location according to the filtering radius.

[0068] Real-time traffic data includes traffic parameters for each road segment within the filtering range, including traffic density and traffic speed. The filtering range is a circular area with the passenger's location as the center and a preset filtering distance as the radius. The filtering distance is preset by those skilled in the art based on actual conditions. Real-time traffic data is obtained through the API of map service providers (such as Gaode Maps, Baidu Maps, Tencent Maps, etc.).

[0069] Traffic density refers to the number of vehicles traveling on different road segments, while traffic speed refers to the average speed of vehicles traveling on different road segments. It should be understood that the higher the traffic density and the slower the traffic speed, the more vehicles there are in the screening area, and the slower the vehicles travel, meaning the screening area is more congested. Therefore, the screening radius should be narrowed to prioritize matching nearby drivers, reducing the driver's travel distance and the passenger's waiting time. Conversely, the lower the traffic density and the faster the traffic speed, the fewer vehicles there are in the screening area, and the faster the vehicles travel. Therefore, the screening radius should be expanded to cover a larger area, which helps to match more drivers, increase the order acceptance rate, and avoid passengers waiting for a long time due to insufficient drivers.

[0070] The steps to define the filter radius include:

[0071] Step S201: Define the iteration threshold a, step size factor b, adjustment factor d, and intensity factor g;

[0072] Step S202: Preset the radius range, which shall be preset by those skilled in the art according to the actual situation;

[0073] Step S203: Construct a population S, which includes m individuals. The position of each individual corresponds one-to-one with the value within the radius range. The initial iteration number t of the population S is 0.

[0074] Step S204: Define the improvement function;

[0075] Step S205: Divide the population S into h subpopulations, each subpopulation containing k individuals, where m = hk;

[0076] Step S206: Calculate the flight probability for each individual and determine the corresponding flight mode;

[0077] Step S207: Update the location of each individual;

[0078] Step S208: Calculate the attraction coefficient for each individual and update the position of each individual again;

[0079] Step S209: Compare the iteration number t with the iteration threshold a; if t≥a, proceed to step S210; if t<a, let t=t+1 and return to step S206.

[0080] Step S210: Calculate the degree of improvement for each individual, and use the value of the individual with the highest degree of improvement as the screening radius.

[0081] In step S201 above, the iteration threshold a, step size factor b, adjustment factor d, and intensity factor g are determined by a person skilled in the art during the historical definition of the screening radius by collecting multiple sets of different real-time traffic data. For each set of real-time traffic data, multiple sets of different definition parameters are set, including the iteration threshold a, step size factor b, adjustment factor d, and intensity factor g. For the same set of real-time traffic data corresponding to different sets of definition parameters, the corresponding screening radius is obtained through the dynamic radius algorithm of steps 1-10, and the corresponding improvement degree is calculated. The definition parameter corresponding to the screening radius with the largest improvement degree is used as the definition parameter corresponding to the real-time traffic data. This process is repeated to obtain the definition parameters corresponding to multiple sets of different real-time traffic data. The average value of the multiple sets of definition parameters (i.e., the average iteration threshold, average step size factor, average adjustment factor, and average intensity factor) is used as the iteration threshold a, step size factor b, adjustment factor d, and intensity factor g defined in step 1.

[0082] In step S203 above, the population S = {X1, X2, ..., X} m}, X m Let m be the m-th individual; the position of each individual in the population S is defined in a one-dimensional search space, the range of which is the radius; the expression for the position of each individual in the population S is: In the formula, Let P be the position of the i-th individual. i Let P be the random coefficient of the i-th individual. i ∈[0,1], i∈[1,m].

[0083] In step S204 above, the expression for the improvement function is: In the formula, f represents the degree of improvement, jd represents the order acceptance time, and ds represents the waiting time. The order acceptance time is the time elapsed from when a passenger places an order to when the driver accepts it, and the waiting time is the time elapsed from when a passenger places an order to when the driver arrives at the passenger's location. The order acceptance time and waiting time are obtained by using real-time traffic data and the numerical values ​​corresponding to individual locations as analysis data. The analysis data is input into a trained first-time model to predict the corresponding order acceptance time, and the analysis data is input into a trained second-time model to predict the corresponding waiting time. Both the first-time model and the second-time model are deep neural network models. Deep neural network models are existing technologies, and the specific training process will not be elaborated on here.

[0084] In step S205 above, the method for dividing population S into h subpopulations includes:

[0085] Calculate the improvement degree for each individual in population S and sort them from largest to smallest. Assign an incremental number to each individual according to the ascending order of sorting, with the sequence number ranging from [1, m]. Based on h subpopulations, perform a modulo operation on the number of each individual to obtain the corresponding sub-number. The expression for the sub-number is: l = u % h; where l is the sub-number, u is the number, and % is the modulo function. If the sub-number is not 0, assign the corresponding individual to the l-th subpopulation; if the sub-number is 0, assign the corresponding individual to the h-th subpopulation.

[0086] For example, population S consists of 3 individuals, which are divided into 3 subpopulations. Since 1%3 = 1, the first individual is assigned to the first subpopulation. Since 2%3 = 2, the second individual is assigned to the second subpopulation. Since 3%3 = 0, the third individual is assigned to the third subpopulation.

[0087] In step S206 above, the expression for the flight probability is: In the formula, Let i be the flight probability corresponding to the i-th individual. f represents the degree of improvement corresponding to the optimal individual. i t Let i be the degree of improvement corresponding to the i-th individual. The worst individual represents the degree of improvement; the best individual is the individual with the highest degree of improvement in population S, and the worst individual is the individual with the lowest degree of improvement in population S.

[0088] Methods for determining an individual's corresponding flight mode include:

[0089] A preset probability threshold is set by a person skilled in the art based on the actual situation. The flight probability of each individual is compared with the probability threshold. If the flight probability is greater than or equal to the probability threshold, the flight mode of the corresponding individual is global flight. If the flight probability is less than the probability threshold, the flight mode of the corresponding individual is local flight.

[0090] In step S207 above, the method for updating the position of each individual includes:

[0091] If the flight mode corresponding to an individual is global flight, the methods for updating the location include:

[0092] In the formula, This represents the position of the i-th individual after the update. b represents the position of the i-th individual before the update. t Let be the step size factor in the t-th iteration. For the optimal position of the individual, d t Let be the adjustment factor in the t-th iteration. C1 is a random number between [0,1]. Let g be the flight speed of the i-th individual. t Let be the intensity factor in the t-th iteration. Let be the flight speed of the i-th individual in the previous iteration;

[0093] The expression for flight speed is: In the formula, C2 is a random number between [0,1].

[0094] If the flight mode corresponding to an individual is local flight, the methods for updating the location include:

[0095] In the formula, Let represent the position of the optimal individual in the subpopulation containing the i-th individual. The optimal individual is the one with the greatest improvement in the subpopulation.

[0096] In step S208 above, the expression for the attraction coefficient is: In the formula, Let be the attraction coefficient between the i'th individual and the i-th individual in the subpopulation containing the i-th individual, and exp be an exponential function. Let be the improvement degree corresponding to the i′ individual in the subpopulation of the i-th individual, i≠i′, i′∈[1,k].

[0097] The methods for updating the location of each individual again include:

[0098] In the formula, This is the position of the i-th individual after the update. This represents the position of the i′th individual in the subpopulation of the i-th individual before the update.

[0099] The dispatch area is a circular area with the passenger's location as the center and the filtering radius as the radius.

[0100] It should be noted that the above individual position update formula guides individual movement by considering factors such as the optimal individual position, its own historical position, flight speed, and random numbers, enabling individuals to converge towards the optimal solution. Specifically, global flight is guided by the globally optimal individual to explore new one-dimensional search spaces and avoid local optima; local flight is guided by the optimal individual in the subpopulation, ensuring that individuals can perform refined searches near local optima and improving local exploration capabilities. When calculating the attraction coefficient, an exponential decay function is used to measure the mutual attraction between individuals. A larger attraction coefficient occurs when the difference in improvement between individuals is small, and a smaller attraction coefficient occurs when the difference in improvement between individuals is large. By introducing an attraction coefficient into the subpopulation, individuals are prevented from converging to the same solution in the one-dimensional search space, thus avoiding a loss of search diversity. The attraction coefficient provides different search paths for each subpopulation, thereby enhancing the algorithm's global search capability and reducing the risk of premature convergence.

[0101] S3: The cloud platform receives driver locations sent by all drivers within the dispatch area.

[0102] The driver's app is the application used by the driver; the driver's location is the geographical coordinates of the driver, which is obtained by the driver's app requesting access to the GPS of a second device, which is the mobile device used by the driver.

[0103] S4: The cloud platform acquires road traffic information, integrates road traffic information, passenger location and all driver locations, calculates the shortest pick-up time from each driver location to the passenger location, and generates a time sorting table.

[0104] Methods for calculating the shortest pick-up time from each driver's location to the passenger's location include:

[0105] Obtain all intersections within the dispatch area through the API of map software (such as Gaode Map, Baidu Map, etc.); based on the passenger location, driver location, and intersections, obtain P pick-up routes from each driver location to the passenger location; calculate the pick-up time corresponding to each pick-up route, compare the pick-up times corresponding to each driver location, and take the shortest pick-up time as the shortest pick-up time for the corresponding driver location.

[0106] The steps to obtain P pick-up routes from each driver's location to each passenger's location include:

[0107] Step S401: Randomly select a driver's position that is not marked as a selected position and mark it as the current position;

[0108] Step S402: Explore all adjacent intersections at the current location, randomly select one of the adjacent intersections as the successor intersection, and mark it as the selected intersection;

[0109] Step S403: Determine whether the passenger's location is adjacent to the next intersection. If so, use the current location, the selected intersection, and the passenger's location as a pick-up route. If not, proceed to step S404.

[0110] Step S404: Explore all adjacent intersections of the successor intersection, randomly select one of the adjacent intersections that is not marked as a selected intersection as the successor intersection, and mark it as a selected intersection;

[0111] Step S405: Determine whether the passenger's location is adjacent to a subsequent intersection. If so, use the current location, the selected intersection, and the passenger's location as a pick-up route. If not, proceed to step S406.

[0112] Step S406: Determine whether all adjacent intersections of the subsequent intersection have been marked as selected intersections. If yes, demark all intersections marked as selected intersections in step S404 and proceed to step S407. If not, return to step S404.

[0113] Step S407: Repeat steps S402 to S406 until P pick-up routes are obtained, then the loop ends and proceed to step S408;

[0114] Step S408: Repeat steps S401 to S407 until all driver positions are marked as the current position. The loop ends, and P pick-up routes from each driver position to the passenger position are obtained.

[0115] The methods for calculating the pick-up time for each pick-up route include:

[0116] Each pick-up route is represented by a set of intersections, with each set corresponding to a specific pick-up route. Two adjacent intersections within each set are grouped into an adjacent set, and the corresponding road segments are identified and marked as pick-up segments. Traffic parameters for each pick-up segment are retrieved from real-time traffic data and marked as pick-up parameters. Road traffic information, including path length and traffic light status data, is obtained via map software API. Path length data includes the path length for each pick-up route. Traffic light status data includes the light status (e.g., red or green) and duration of each intersection along the pick-up route. Different numerical labels are assigned to different light statuses and marked as light labels. The light status data is then replaced with the corresponding light labels. The pick-up parameters, path length, and traffic light status data for each pick-up route are treated as a set of computational data, each corresponding to a specific pick-up route. Each set of computational data is input into a trained time prediction model, which is a deep neural network model, to predict the corresponding pick-up time.

[0117] Methods for generating time-sorted tables include:

[0118] Sort all the shortest pick-up times from shortest to longest to generate a time sorting table.

[0119] S5: The cloud platform sends the order to the driver with the shortest pick-up time. If the driver does not accept the order, the cloud platform will assign orders to the drivers in order according to the time sorting table.

[0120] If no driver accepts the order, the driver's terminal is assigned an order in ascending order according to the time sorting table. If no driver accepts the order, the filtering radius is multiplied by a preset expansion coefficient to obtain the expansion radius. Based on the expansion radius, the passenger's location is re-generated and the order is reassigned. The expansion coefficient is preset by those skilled in the art according to the actual situation.

[0121] This embodiment acquires passenger location and real-time traffic data, and uses a dynamic radius algorithm to adaptively generate dispatch areas, thereby dynamically adjusting the dispatch range and improving dispatch accuracy. It calculates the shortest pick-up time from each driver's location to the passenger's location in real time, prioritizing dispatches to the driver closest to the passenger, thus reducing passenger waiting time and driver empty-running rate, and effectively improving user experience. It fully utilizes cloud computing and big data analytics to achieve intelligent optimization and dynamic adjustment of order dispatch, improving overall operational efficiency and adapting to travel demands in the context of the sharing economy.

[0122] Example 2

[0123] This application also provides an electronic device. The electronic device may include one or more processors and one or more memories. The memories store computer-readable code, which, when executed by the one or more processors, can perform an order management method based on a cloud computing service as described above.

[0124] The method or system according to the embodiments of this application can also be implemented using the architecture of the electronic device shown in this application. The electronic device may include a bus, one or more CPUs, ROM, RAM, a communication port connected to a network, input / output, a hard disk, etc. The storage device in the electronic device, such as ROM or hard disk, may store an order management method based on cloud computing services provided in this application. Furthermore, the electronic device may also include a user interface. Of course, the architecture shown in this application is merely exemplary; when implementing different devices, one or more components of the electronic device shown in this application may be omitted according to actual needs.

[0125] Example 3

[0126] Please refer to the accompanying drawings. One embodiment of this application discloses a computer-readable storage medium. The computer-readable storage medium stores computer-readable instructions. When executed by a processor, the computer-readable instructions can perform an order management method based on a cloud computing service, as described in the above figures, according to an embodiment of this application. The storage medium includes, but is not limited to, volatile memory and / or non-volatile memory. Volatile memory may include, for example, random access memory (RAM) and cache memory. Non-volatile memory may include, for example, read-only memory (ROM), hard disk, flash memory, etc.

[0127] Furthermore, according to embodiments of this application, the processes described in the above-referenced flowcharts can be implemented as computer software programs. For example, this application provides a non-transitory machine-readable storage medium storing machine-readable instructions that can be executed by a processor to perform instructions corresponding to the method steps provided in this application, such as an order management method based on cloud computing services. When this computer program is executed by a central processing unit (CPU), it performs the functions defined in the method of this application.

[0128] The above description is merely a specific embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Any variations or substitutions that can be easily conceived by those skilled in the art within the technical scope disclosed in the present invention should be included within the scope of protection of the present invention. Therefore, the scope of protection of the present invention should be determined by the scope of the claims.

[0129] In conclusion, the above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the protection scope of the present invention.

Claims

1. An order management method based on cloud computing services, characterized in that, Applications in cloud platforms include: S1: The cloud platform receives the passenger's location sent by the passenger terminal; S2: The cloud platform acquires real-time traffic data, defines a filtering radius using a dynamic radius algorithm based on the real-time traffic data, and generates a dispatch area based on the passenger's location based on the filtering radius; S3: The cloud platform receives the driver locations sent by all drivers within the dispatch area; S4: The cloud platform acquires road traffic information, integrates road traffic information, passenger location and all driver locations, calculates the shortest pick-up time from each driver location to the passenger location, and generates a time sorting table; S5: The cloud platform sends the order to the driver with the shortest pick-up time. If the driver does not accept the order, the cloud platform will assign orders to the drivers in order according to the time sorting table.

2. The order management method based on cloud computing services according to claim 1, characterized in that, The passenger location is the geographical coordinate of the passenger, and the driver location is the geographical coordinate of the driver. Real-time traffic data includes traffic parameters for each road segment within the filtering range, including traffic density and traffic speed; the filtering range is a circular area with the passenger's location as the center and a preset filtering distance as the radius; The steps to define the filter radius include: Step S201: Define the iteration threshold a, step size factor b, adjustment factor d, and intensity factor g; Step S202: Preset radius range; Step S203: Construct a population S, which includes m individuals. The position of each individual corresponds one-to-one with the value within the radius range. The initial iteration number t of the population S is 0. Step S204: Define the improvement function; Step S205: Divide the population S into h subpopulations, each subpopulation containing k individuals, where m = hk; Step S206: Calculate the flight probability for each individual and determine the corresponding flight mode; Step S207: Update the location of each individual; Step S208: Calculate the attraction coefficient for each individual and update the position of each individual again; Step S209: Compare the iteration number t with the iteration threshold a; if t≥a, proceed to step S210; if t<a, let t=t+1 and return to step S206. Step S210: Calculate the degree of improvement for each individual, and use the value of the individual with the highest degree of improvement as the screening radius.

3. The order management method based on cloud computing services according to claim 2, characterized in that, In step S203, the population S = {X1, X2, ..., X} m }, X m Let m be the m-th individual; the position of each individual in the population S is defined in a one-dimensional search space, the range of which is the radius; the expression for the position of each individual in the population S is: In the formula, Let P be the position of the i-th individual. i Let P be the random coefficient of the i-th individual. i ∈[0,1], i∈[1,m]; In step S204, the expression for the improvement degree function is: In the formula, f represents the degree of improvement, jd represents the order acceptance time, and ds represents the waiting time. The order acceptance time is the time elapsed from when a passenger places an order to when the driver accepts the order, and the waiting time is the time elapsed from when a passenger places an order to when the driver arrives at the passenger's location. The order acceptance time and the waiting time are obtained by using real-time traffic data and the numerical values ​​corresponding to individual locations as analysis data. The analysis data is input into a trained first-time model to predict the corresponding order acceptance time, and the analysis data is input into a trained second-time model to predict the corresponding waiting time. Both the first-time model and the second-time model are deep neural network models.

4. The order management method based on cloud computing services according to claim 3, characterized in that, In step S205, the method for dividing the population S into h subpopulations includes: Calculate the improvement degree for each individual in population S and sort them from largest to smallest. Assign an incremental number to each individual according to the ascending order, with the sequence number ranging from [1, m]. For each of the h subpopulations, perform a modulo operation on the number of each individual to obtain the corresponding sub-number. The expression for the sub-number is: l = u % h; where l is the sub-number, u is the number, and % is the modulo function. If the sub-number is not 0, the corresponding individual is assigned to the l-th subpopulation; if the sub-number is 0, the corresponding individual is assigned to the h-th subpopulation. In step S206, the expression for the flight probability is: In the formula, Let i be the flight probability corresponding to the i-th individual. f represents the degree of improvement corresponding to the optimal individual. i t Let i be the degree of improvement corresponding to the i-th individual. The worst individual is the one with the greatest improvement in population S; the best individual is the one with the greatest improvement in population S, and the worst individual is the one with the least improvement in population S. Methods for determining an individual's corresponding flight mode include: A preset probability threshold is set, and the flight probability of each individual is compared with the probability threshold. If the flight probability is greater than or equal to the probability threshold, the flight mode of the corresponding individual is global flight; if the flight probability is less than the probability threshold, the flight mode of the corresponding individual is local flight.

5. The order management method based on cloud computing services according to claim 4, characterized in that, In step S207, the method for updating the position of each individual includes: If the flight mode corresponding to an individual is global flight, the methods for updating the location include: In the formula, This represents the position of the i-th individual after the update. b represents the position of the i-th individual before the update. t Let be the step size factor in the t-th iteration. For the optimal position of the individual, d t Let be the adjustment factor in the t-th iteration. C1 is a random number between [0,1]. Let g be the flight speed of the i-th individual. t Let be the intensity factor in the t-th iteration. Let be the flight speed of the i-th individual in the previous iteration; The expression for flight speed is: In the formula, C2 is a random number between [0,1]. If the flight mode corresponding to an individual is local flight, the methods for updating the location include: In the formula, Let represent the position of the optimal individual in the subpopulation containing the i-th individual. The optimal individual is the one with the greatest improvement in the subpopulation.

6. The order management method based on cloud computing services according to claim 5, characterized in that, In step S208, the expression for the attraction coefficient is: In the formula, Let be the attraction coefficient between the i'th individual and the i-th individual in the subpopulation containing the i-th individual, and exp be an exponential function. Let be the degree of improvement corresponding to the i′ individual in the subpopulation of the i-th individual, i≠i′, i′∈[1,k]; The methods for updating the location of each individual again include: In the formula, This is the position of the i-th individual after the update. To update the position of the i′th individual in the subpopulation of the i-th individual before the update; The dispatch area is a circular area with the passenger's location as the center and the filtering radius as the radius.

7. The order management method based on cloud computing services according to claim 6, characterized in that, The method for calculating the shortest pick-up time from each driver's location to the passenger's location includes: Get all intersections within the dispatch area; based on the passenger location, driver location, and intersection, obtain P pick-up routes from each driver location to the passenger location; calculate the pick-up time corresponding to each pick-up route, compare the pick-up times corresponding to each driver location, and take the shortest pick-up time as the shortest pick-up time for the corresponding driver location.

8. The order management method based on cloud computing services according to claim 7, characterized in that, The steps for obtaining P pick-up routes from each driver's location to each passenger's location include: Step S401: Randomly select a driver's position that is not marked as a selected position and mark it as the current position; Step S402: Explore all adjacent intersections at the current location, randomly select one of the adjacent intersections as the successor intersection, and mark it as the selected intersection; Step S403: Determine whether the passenger's location is adjacent to the next intersection. If so, use the current location, the selected intersection, and the passenger's location as a pick-up route. If not, proceed to step S404. Step S404: Explore all adjacent intersections of the successor intersection, randomly select one of the adjacent intersections that is not marked as a selected intersection as the successor intersection, and mark it as a selected intersection; Step S405: Determine whether the passenger's location is adjacent to a subsequent intersection. If so, use the current location, the selected intersection, and the passenger's location as a pick-up route. If not, proceed to step S406. Step S406: Determine whether all adjacent intersections of the subsequent intersection have been marked as selected intersections. If yes, demark all intersections marked as selected intersections in step S404 and proceed to step S407. If not, return to step S404. Step S407: Repeat steps S402 to S406 until P pick-up routes are obtained, then the loop ends and proceed to step S408; Step S408: Repeat steps S401 to S407 until all driver positions are marked as the current position. The loop ends, and P pick-up routes from each driver position to the passenger position are obtained.

9. The order management method based on cloud computing services according to claim 8, characterized in that, The method for calculating the pick-up time for each pick-up route includes: Each pick-up route is represented by a set of intersections, with each set corresponding to a specific pick-up route. Two adjacent intersections within each set are grouped into an adjacent set, and the corresponding road segments are identified and marked as pick-up segments. Traffic parameters for each pick-up segment are retrieved from real-time traffic data and marked as pick-up parameters. Road traffic information, including path length and traffic light status data, is acquired. Path length data includes the path length for each pick-up route. Traffic light status data includes the light status and duration at each intersection along each pick-up route. Different numerical labels are assigned to different light statuses and marked as light labels. The light status data is then replaced with the corresponding light labels. The pick-up parameters, path length, and traffic light status data for each pick-up route are treated as a set of computational data, with each set corresponding to a specific pick-up route. Each set of computational data is input into a trained time prediction model, which is a deep neural network model, to predict the corresponding pick-up time. Methods for generating time-sorted tables include: Sort all the shortest pick-up times from shortest to longest to generate a time sorting table.

10. The order management method based on cloud computing services according to claim 9, characterized in that, If no driver accepts the order, the driver's terminal will be assigned orders in ascending order according to the time sorting table. If no driver accepts the order, the filtering radius will be multiplied by a preset expansion coefficient to obtain the expansion radius. Based on the expansion radius, the order assignment area will be generated again for the passenger's location, and the order will be reassigned.

Images