Optimization method formed through secondary feedback of rider

Abstract

The invention relates to the technical field of distribution management, in particular to an optimization method formed through secondary feedback of a rider, and the method comprises the steps: S1, building a networking platform, and collecting various information of the rider in real time; s2, a feedback mechanism is established, and a rider feeds back important information encountered in the distribution process in real time; s3, collecting the work data of the rider, and generating a personalized work capability evaluation model; s4, the networked map software collects road information, and establishes a route prediction mechanism in combination with feedback data and historical data; s5, monitoring distribution information of the rider in real time, and providing a processing scheme in case of an emergency; and S6, obtaining distribution information, and learning and optimizing an evaluation model and route planning of the rider. The method is beneficial for greatly improving the flexibility and response speed of the platform to cope with emergencies, ensuring dynamic optimization of the distribution plan of the rider, improving the consumer experience, enhancing the participation sense of the rider, and improving the working enthusiasm and initiative of the rider.

Description

Technical Field

[0001] The present invention relates to the technical field of delivery management, and in particular to an optimization method formed through secondary feedback from riders. Background Art

[0002] The food delivery platform is a service platform based on the Internet and mobile applications, which connects consumers, catering merchants and riders to provide users with food delivery services. The platform's algorithm system assigns orders to riders based on factors such as geographic location and order volume. After the riders pick up the food from the catering merchants, they deliver the food to the consumers.

[0003] The patent application number is CN202210186786.5, and it is recorded in the specification that "the present invention designs a method for evaluating the reasonable delivery time of food delivery riders by comprehensively considering the internal and external influencing factors of food delivery riders. In order to standardize the reasonable delivery time and monitor the rationality of the food delivery platform's delivery time specifications, it not only takes into account the external weather and traffic influencing factors, but also takes into account the rider's single load and the day's workload factors, reasonably determines the duration, and effectively guarantees labor safety, so that riders do not have to ride fast, and consumers also have a reasonable time to place orders and wait for meals, thereby solving unnecessary delivery conflicts between the two parties." Although the above method has There are methods that take into account external and internal factors such as weather, traffic, and rider load to solve the problem of delivery time. However, the above methods rely on a large amount of real-time data, such as weather, traffic, and traffic light duration. Ensuring the accuracy and timeliness of the data is a challenge for the platform, which may cause the platform's computing power to be overloaded. When data is delayed or inaccurate, it will affect the delivery effect. In addition, although the above methods take into account the rider's load and the workload of the day, they ignore the differences between different riders. Factors such as different riders' riding speed, proficiency, and ability to adapt to different weather and traffic conditions will affect the delivery effect.

[0004] To sum up, developing an optimization method based on secondary feedback from riders is still a key issue that needs to be urgently addressed in the field of food delivery technology. Summary of the invention

[0005] In order to solve the problem in the prior art that the above method has comprehensive consideration of external and internal factors such as weather, traffic, rider load, etc., so as to solve the problem of delivery time, but the above method relies on a large amount of real-time data, such as weather, traffic, traffic light duration, etc., and ensuring the accuracy and timeliness of the data is a challenge to the platform, which leads to overload of the platform's computing power. When data is delayed or inaccurate, it will affect the delivery effect. In addition, although the above method takes into account the rider's load and the workload of the day, it ignores the differences between different riders. Factors such as the riding speed, proficiency, and adaptability to different weather and traffic conditions of different riders will affect the delivery effect. The present invention provides an optimization method formed through secondary feedback from riders.

[0006] To achieve the above object, the present invention provides the following technical solutions:

[0007] The present invention provides an optimization method formed by secondary feedback from the rider, comprising:

[0008] S1. Build an Internet-connected platform to collect various information of riders in real time;

[0009] S2. Establish a feedback mechanism so that riders can provide real-time feedback on important information encountered during the delivery process;

[0010] S3. Collect riders’ work data and generate a personalized work ability assessment model;

[0011] S4. Networked map software collects road information and combines feedback data with historical data to establish a route prediction mechanism;

[0012] S5. Real-time monitoring of rider delivery information and providing solutions in case of emergencies;

[0013] S6. Obtain delivery information, learn and optimize the rider’s evaluation model and route planning.

[0014] Furthermore, in step S1, the method of building a network platform to collect various information of riders in real time is:

[0015] Build an Internet-connected platform that connects food delivery platforms, riders and consumers, and then use the platform to obtain real-time data on rider delivery orders. The data includes but is not limited to order generation time, delivery distance, restaurant location and delivery address. At the same time, the Internet-connected platform will obtain the rider's daily workload and load information in real time during the delivery process, and then upload the rider's real-time location, driving speed and road condition information to the platform through IoT devices.

[0016] Furthermore, in step S2, a feedback mechanism is established, and the method for the rider to provide real-time feedback on important information encountered during the delivery process is as follows:

[0017] Riders use applications on their mobile internet-connected devices to provide feedback during the delivery process on issues including but not limited to traffic jams, road closures, long wait times for food, and delays in elevator use. Feedback can be provided by clicking on preset options or through voice input.

[0018] Furthermore, in step S2, a feedback mechanism is established, and the method for the rider to provide real-time feedback on important information encountered during the delivery process is as follows:

[0019] When a rider encounters an emergency, the feedback delay time is expressed as: S T (n) = x n ×F n , where S T (n) represents the delay time, F n represents the impact time of the nth feedback event, χ n It is a coefficient set according to the feedback type. When a rider has multiple feedbacks in the same delivery process, the cumulative feedback delay time is expressed as: Where m is the total number of events reported by the rider, S T is the accumulated feedback delay time.

[0020] Furthermore, in step S3, the method of collecting the rider's work data and generating a personalized work ability evaluation model is as follows:

[0021] The platform collects the riders' historical work performance, including but not limited to average delivery speed, ability to cope with different weather and traffic conditions, and work efficiency during single and multiple deliveries. The platform combines these historical data with current order data, including but not limited to delivery distance, weather, and traffic routes, and uses machine learning algorithms to generate personalized delivery time recommendations, creating a personalized work ability assessment model for each rider.

[0022] Furthermore, in step S3, the method of collecting the rider's work data and generating a personalized work ability evaluation model is as follows:

[0023] Taking into account the rider's historical data, the day's workload, and environmental factors, the formula for evaluating the rider's delivery time is: q =A base ×(1-ζ1×M-ζ2×S T -ζ3×I), where Α q is the delivery speed of the rider, A base is the basic riding speed of the rider, which is taken from his historical data. M is the cumulative workload of the rider on that day, including but not limited to the number of delivered orders. S Tis the influencing factor of the current weather conditions, including but not limited to deceleration in rainy days, waiting for traffic lights and road congestion. I is the load of a single delivery, including but not limited to the weight of the order or the number of deliveries. ζ1, ζ2 and ζ3 are the corresponding influencing coefficients, which can be obtained through regression analysis of historical data. The delivery time of the rider will be affected by the speed and delivery distance. The expression formula of the rider's personalized delivery time is: Where Q L is the personalized delivery time, X is the delivery distance, A q is the rider’s personalized speed, G xh Other fixed time consumption includes but is not limited to parking, going up and down elevators, and handing over items. When multiple influencing factors need to be considered comprehensively, the formula for expressing the rider's personalized delivery time can be expanded to: Among them G jt It is the influencing factor of traffic conditions and is obtained through traffic data.

[0024] Furthermore, in step S4, the networked map software collects road information, and then combines the feedback data and historical data to establish a route prediction mechanism as follows:

[0025] After receiving the feedback data, the platform immediately calculates the impact of these emergencies on the delivery time. The platform adjusts the estimated delivery time of the current order based on the feedback data, and notifies the user and the food delivery platform for secondary feedback when necessary. The platform can also optimize the traffic route selection based on repeated emergencies to avoid similar situations from happening again. The expression formula for the estimated delivery time combined with the feedback delay is: S fkh =S zcsj +S T , where S zcsj is the initial delivery time evaluation result without considering the feedback factor, S T is the accumulated delay introduced by rider feedback.

[0026] Furthermore, in step S4, the networked map software collects road information, and then combines the feedback data and historical data to establish a route prediction mechanism as follows:

[0027] The platform's networked map software takes into account the current situation, including but not limited to traffic conditions, weather conditions, and road construction, and dynamically optimizes the rider's delivery route. When the rider reports in real time that there is traffic congestion or a blocked road ahead, the platform's networked map software immediately gives the best route again and recommends detours to the rider to reduce delivery delays. In addition, the platform can predict traffic conditions in specific time periods and locations by analyzing historical traffic data from the map software, and optimize the rider's delivery route in advance to avoid congestion during peak hours. The platform's optimized route expression formula is: E = ZXT d (B), where B is the Bth alternative route, Td (B) depends on the conditions of each route, including but not limited to traffic conditions, road length, and weather. During the delivery process, if there is an emergency in the road conditions, the platform will adjust the delivery route in real time based on the feedback. The expression formula is: S new =ZX[T d (B)+T d ], select route B with the least delay after adjustment, and use the historical traffic data to predict the traffic conditions in a certain period of time in the future: S tf (t) = ∫yzt(t, X, G, J), where ∫yzt is a traffic prediction function based on historical data modeling, t is the estimated time, X is the route distance, G is the weekday or weekend influencing factor, and J is the time period including but not limited to peak hours and non-peak hours. The platform uses a comprehensive feedback mechanism, personalized evaluation and route prediction to obtain the final dynamic optimization of delivery time expression formula: Among them A T (B) is the rider’s personalized speed on route t, T d (B) is the delay time of the route, including but not limited to the influence of traffic, weather, traffic lights and merchant’s food delivery feedback.

[0028] Furthermore, in step S5, the method of real-time monitoring of the rider's delivery information and providing a solution in the face of emergencies is as follows:

[0029] Based on the data from the riders’ real-time feedback, including but not limited to traffic jams and bad weather, the platform automatically sends delay notification information to the food delivery platform and users when the estimated delivery time is not within the predetermined standard. Users can view the latest estimated delivery time through the platform and choose to continue waiting or cancel the order based on the delay. The platform can automatically split the order based on the riders’ feedback and assign part of the delivery tasks to other nearby riders to reduce the workload of the current riders. At the same time, the platform adjusts the delivery order of multiple orders based on the riders’ secondary feedback data, giving priority to orders that are closer or with tight estimated time. When the platform finds a delay, the expression formula for determining which orders need to be prioritized is:

[0030] Where Z p is the order priority, D dly is the delay time of the order, X is the delivery distance of the order, J p The order value includes but is not limited to orders with larger amounts, which have higher priority. κ1, κ2, and κ3 are weight coefficients. m is the current number of orders for the rider. When the platform finds that the rider cannot complete the task, the expression formula for reallocation is: Among them I e is the current rider's load, X eis the distance between the rider and the order destination, τ1, τ2 are weight coefficients, y is the total number of currently available riders, and the platform automatically updates the estimated delivery time as follows: U x =U yj +U dly +U aut , where U x is the estimated delivery time after automatic update. yj is the current estimated delivery time, U dly is the delay time estimated based on real-time feedback, U aut It is the additional time after the system reallocates or adjusts the route. The platform splits an order into multiple parts according to the expression formula: Among them, H i is the cost function of order splitting, S q (n) is the delivery time of the order part, and q is the number of parts after the order is split.

[0031] Furthermore, in step S6, the method of real-time monitoring of the rider's delivery information and providing a solution in the face of emergencies is as follows:

[0032] Every feedback from the riders and the delivery results of each order, including but not limited to whether it is delivered on time and the situations encountered during the delivery process, will be recorded and analyzed by the platform. The platform uses machine learning technology to conduct in-depth analysis of all feedback data and delivery data, find out the deficiencies of the current evaluation model and rules, and regularly update and optimize the rules. The weight coefficients of factors affecting delivery time are continuously optimized through the gradient descent method. The expression formula is: S q =μ0+μ1G+μ2M t +μ3Z+μ4J tfc , where S q is the predicted delivery time, G is the rider workload, M t is the weather condition, Z is the rider load, J tfc is the traffic condition, μ0, μ1, μ2, μ3, μ4 are coefficients, and the actual delivery time S sj and the predicted delivery time S q The difference is used to update the coefficient, and the expression formula is: Where α is the learning rate, is the gradient of the loss function. In order to combine the influence of historical data and the latest feedback data, the platform dynamically adjusts the prediction time through the weighted average method. The expression formula is: in is the predicted time based on historical data, is the prediction time based on feedback data, l is the weight coefficient, which is used to balance the impact of historical data and real-time feedback. It is used to continuously adjust the hyperparameters of the model through Bayesian optimization. The expression formula is: Where f(θ) is the loss function of the model, and θ is a hyperparameter including but not limited to learning rate and regularization parameter.

[0033] Beneficial Effects

[0034] Compared with the known public technology, the technical solution provided by the present invention has the following advantages:

[0035] Beneficial effects:

[0036] The present invention is beneficial to greatly improve the flexibility and response speed of the platform in dealing with emergencies, ensure the dynamic optimization of the rider's delivery plan, improve consumer experience, enhance rider participation, and improve rider's work enthusiasm and initiative.

[0037] The present invention generates a personalized work ability evaluation model to assign more suitable orders to each rider, thereby improving the overall delivery efficiency. The continuous accumulation of riders' work data can help riders identify shortcomings in their work, thereby providing them with optimization suggestions and further improving the riders' overall work performance. At the same time, the platform can make fair and accurate rider performance evaluations based on data, which helps to formulate a reward mechanism to motivate riders with outstanding performance.

[0038] The present invention predicts the road conditions that may occur during the rider's work and reminds the rider to adjust the route in time, ensuring that the rider chooses the optimal delivery route and reduces time delays. This is conducive to improving consumers' trust in delivery services, while helping riders reduce driving time and reduce delivery difficulty. By intelligently recommending processing solutions, it is conducive to helping riders deal with difficulties in the delivery process and improving overall delivery efficiency.

[0039] The present invention optimizes the rider's personalized work ability assessment model through machine learning technology, and obtains the best delivery mode based on the data feedback after each delivery, so as to further improve the processing method of future orders, which is conducive to optimizing the delivery strategy and route planning, ensuring that each delivery is completed in the shortest possible time, reducing the burden on the rider and improving the adaptability of the platform. BRIEF DESCRIPTION OF THE DRAWINGS

[0040] Figure 1 The present invention is a flow chart of an optimization method formed through secondary feedback from riders. DETAILED DESCRIPTION

[0041] In order to enable those skilled in the art to better understand the scheme of the present invention, the technical scheme in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work should fall within the scope of protection of the present invention.

[0042] It should be noted that the terms "first", "second", etc. in the specification and claims of the present invention and the above-mentioned drawings are used to distinguish similar objects, and are not necessarily used to describe a specific order or sequence. It should be understood that the data used in this way can be interchanged where appropriate, so that the embodiments of the present invention described herein can be implemented in an order other than those illustrated or described herein. In addition, the terms "including" and "having" and any variations thereof are intended to cover non-exclusive inclusions, for example, a process, method, system, product or device that includes a series of steps or units is not necessarily limited to those steps or units that are clearly listed, but may include other steps or units that are not clearly listed or inherent to these processes, methods, products or devices.

[0043] The present invention is further described in detail below in conjunction with the accompanying drawings:

[0044] Example:

[0045] like Figure 1 As shown, the present invention provides an optimization method formed by secondary feedback from the rider, including:

[0046] S1. Build an Internet-connected platform to collect various information of riders in real time;

[0047] Furthermore, in step S1, the method of building a network platform to collect various information of riders in real time is:

[0048] Build an Internet-connected platform that connects food delivery platforms, riders and consumers, and then use the platform to obtain real-time data on rider delivery orders. The data includes but is not limited to order generation time, delivery distance, restaurant location and delivery address. At the same time, the Internet-connected platform will obtain the rider's daily workload and load information in real time during the delivery process, and then upload the rider's real-time location, driving speed, and road condition information to the platform through IoT devices to avoid delays in rider delivery work caused by information lags.

[0049] In this embodiment, the platform collects real-time data of riders, food delivery platforms, and consumers, and then integrates and stores this information in a back-end database for subsequent analysis and processing, thereby performing information interaction and ensuring data update and consistency.

[0050] S2. Establish a feedback mechanism so that riders can provide real-time feedback on important information encountered during the delivery process;

[0051] Furthermore, in step S2, a feedback mechanism is established, and the method for the rider to provide real-time feedback on important information encountered during the delivery process is as follows:

[0052] Riders use applications on their mobile internet-connected devices to provide feedback during the delivery process on issues including but not limited to traffic jams, road closures, long wait times for food, and delays in elevator use. Feedback can be provided by clicking on preset options or through voice input.

[0053] Furthermore, in step S2, a feedback mechanism is established, and the method for the rider to provide real-time feedback on important information encountered during the delivery process is as follows:

[0054] When a rider encounters an emergency, the feedback delay time is expressed as: S T (n) = x n ×F n , where S T (n) represents the delay time, F n represents the impact time of the nth feedback event, χ n It is a coefficient set according to the feedback type. When a rider has multiple feedbacks in the same delivery process, the cumulative feedback delay time is expressed as: Where m is the total number of events reported by the rider, S T is the accumulated feedback delay time.

[0055] In this embodiment, a feedback mechanism is used to collect emergencies in the delivery process in real time, so that the delivery route or the estimated arrival time can be adjusted according to the feedback data. At the same time, the platform sets priorities according to the severity of the feedback event. For high-level events, the platform will prioritize computing power and provide reasonable processing suggestions to riders through data analysis, which is conducive to greatly improving the platform's flexibility and response speed in dealing with emergencies, ensuring dynamic optimization of riders' delivery plans, improving consumer experience, enhancing rider participation, and improving riders' work enthusiasm and initiative.

[0056] S3. Collect riders’ work data and generate a personalized work ability assessment model;

[0057] Furthermore, in step S3, the method of collecting the rider's work data and generating a personalized work ability evaluation model is as follows:

[0058] The platform collects the riders' historical work performance, including but not limited to average delivery speed, ability to cope with different weather and traffic conditions, and work efficiency during single and multiple deliveries. The platform combines these historical data with current order data, including but not limited to delivery distance, weather, and traffic routes, and uses machine learning algorithms to generate personalized delivery time recommendations, creating a personalized work ability assessment model for each rider.

[0059] Furthermore, in step S3, the method of collecting the rider's work data and generating a personalized work ability evaluation model is as follows:

[0060] Taking into account the rider's historical data, the day's workload, and environmental factors, the formula for evaluating the rider's delivery time is: q =A base ×(1-ζ1×M-ζ2×S T -ζ3×I), where Α q is the delivery speed of the rider, A base is the basic riding speed of the rider, which is taken from his historical data. M is the cumulative workload of the rider on that day, including but not limited to the number of delivered orders. S T is the influencing factor of the current weather conditions, including but not limited to deceleration in rainy days, waiting for traffic lights and road congestion. I is the load of a single delivery, including but not limited to the weight of the order or the number of deliveries. ζ1, ζ2 and ζ3 are the corresponding influencing coefficients, which can be obtained through regression analysis of historical data. The delivery time of the rider will be affected by the speed and delivery distance. The expression formula of the rider's personalized delivery time is: Where Q L is the personalized delivery time, X is the delivery distance, A q is the rider’s personalized speed, G xh Other fixed time consumption includes but is not limited to parking, going up and down elevators, and handing over items. When multiple influencing factors need to be considered comprehensively, the formula for expressing the rider's personalized delivery time can be expanded to: Among them G jt It is the influencing factor of traffic conditions and is obtained through traffic data.

[0061] In this embodiment, the platform accumulates the riders' work performance data through long-term collection of the riders' work data. The platform generates a personalized work ability assessment model based on the riders' work performance data, thereby assigning more suitable orders to each rider and improving the overall delivery efficiency. The continuous accumulation of riders' work data can help riders identify shortcomings in their work, thereby providing them with optimization suggestions and further improving the riders' overall work performance. At the same time, the platform can make fair and accurate rider performance evaluations based on data, which helps to formulate a reward mechanism to motivate riders with outstanding performance.

[0062] S4. Networked map software collects road information and combines feedback data with historical data to establish a route prediction mechanism;

[0063] Furthermore, in step S4, the networked map software collects road information, and then combines the feedback data and historical data to establish a route prediction mechanism as follows:

[0064] After receiving the feedback data, the platform immediately calculates the impact of these emergencies on the delivery time. The platform adjusts the estimated delivery time of the current order based on the feedback data, and notifies the user and the food delivery platform for secondary feedback when necessary. The platform can also optimize the traffic route selection based on repeated emergencies to avoid similar situations from happening again. The expression formula for the estimated delivery time combined with the feedback delay is: S fkh =S zcsj +S T , where S zcsj is the initial delivery time evaluation result without considering the feedback factor, S T is the accumulated delay introduced by rider feedback.

[0065] Furthermore, in step S4, the networked map software collects road information, and then combines the feedback data and historical data to establish a route prediction mechanism as follows:

[0066] The platform's networked map software takes into account the current situation, including but not limited to traffic conditions, weather conditions, and road construction, and dynamically optimizes the rider's delivery route. When the rider reports in real time that there is traffic congestion or a blocked road ahead, the platform's networked map software immediately gives the best route again and recommends detours to the rider to reduce delivery delays. In addition, the platform can predict traffic conditions in specific time periods and locations by analyzing historical traffic data from the map software, and optimize the rider's delivery route in advance to avoid congestion during peak hours. The platform's optimized route expression formula is: E = ZXT d (B), where B is the Bth alternative route, T d (B) depends on the conditions of each route, including but not limited to traffic conditions, road length, and weather. During the delivery process, if there is an emergency in the road conditions, the platform will adjust the delivery route in real time based on the feedback. The expression formula is: S new =ZX[T d (B)+T d ], select route B with the least delay after adjustment, and use the historical traffic data to predict the traffic conditions in a certain period of time in the future: S tf(t) = ∫yzt(t, X, G, J), where ∫yzt is a traffic prediction function based on historical data modeling, t is the estimated time, X is the route distance, G is the weekday or weekend influencing factor, and J is the time period including but not limited to peak hours and non-peak hours. The platform uses a comprehensive feedback mechanism, personalized evaluation and route prediction to obtain the final dynamic optimization of delivery time expression formula: Among them A T (B) is the rider’s personalized speed on route t, T d (B) is the delay time of the route, including but not limited to the influence of traffic, weather, traffic lights and merchant’s food delivery feedback.

[0067] In this embodiment, the platform obtains road information in real time through the online map API. Combining the secondary real-time feedback data from the riders, historical delivery data, and current road information, the platform establishes route prediction data to predict road conditions that may arise during the riders' work, including but not limited to traffic lights, weather conditions, and road construction. The platform reminds the riders to adjust their routes in a timely manner to ensure that the riders choose the optimal delivery route and reduce time delays, which is conducive to enhancing consumers' trust in delivery services while helping riders reduce driving time and reduce delivery difficulty.

[0068] S5. Real-time monitoring of rider delivery information and providing solutions in case of emergencies;

[0069] Furthermore, in step S5, the method of real-time monitoring of the rider's delivery information and providing a solution in the face of emergencies is as follows:

[0070] Based on the data from the riders’ real-time feedback, including but not limited to traffic jams and bad weather, the platform automatically sends delay notification information to the food delivery platform and users when the estimated delivery time is not within the predetermined standard. Users can view the latest estimated delivery time through the platform and choose to continue waiting or cancel the order based on the delay. The platform can automatically split the order based on the riders’ feedback and assign part of the delivery tasks to other nearby riders to reduce the workload of the current riders. At the same time, the platform adjusts the delivery order of multiple orders based on the riders’ secondary feedback data, giving priority to orders that are closer or with tight estimated time. When the platform finds a delay, the expression formula for determining which orders need to be prioritized is:

[0071] Where Z p is the order priority, D dly is the delay time of the order, X is the delivery distance of the order, J pThe order value includes but is not limited to orders with larger amounts, which have higher priority. κ1, κ2, and κ3 are weight coefficients. m is the current number of orders for the rider. When the platform finds that the rider cannot complete the task, the expression formula for reallocation is: Among them I e is the current rider's load, X e is the distance between the rider and the order destination, τ1, τ2 are weight coefficients, y is the total number of currently available riders, and the platform automatically updates the estimated delivery time as follows: U x =U yj +U dly +U aut , where U x is the estimated delivery time after automatic update. yj is the current estimated delivery time, U dly is the delay time estimated based on real-time feedback, U aut It is the additional time after the system reallocates or adjusts the route. The platform splits an order into multiple parts according to the expression formula: Among them, H i is the cost function of order splitting, S q (n) is the delivery time of the order part, and q is the number of parts after the order is split.

[0072] In this embodiment, the platform continuously monitors the rider's delivery information to determine the rider's delivery status. If an abnormality is detected, including but not limited to long-term stagnation and deviation from the scheduled route, the platform will send an alarm message to the rider and require the rider to provide feedback. If the rider does not provide feedback for a long time, the rider is defined as an abnormal rider. The platform intelligently recommends a processing plan, including but not limited to sending a search message to other nearby riders, and obtaining secondary feedback from other riders to understand the information of the abnormal rider, and then assist in handling abnormal delivery orders, which is conducive to helping riders deal with difficulties in the delivery process and improving overall delivery efficiency.

[0073] S6. Obtain delivery information, learn and optimize the rider’s evaluation model and route planning;

[0074] Furthermore, in step S6, the method of real-time monitoring of the rider's delivery information and providing a solution in the face of emergencies is as follows:

[0075] Every feedback from the riders and the delivery results of each order, including but not limited to whether it is delivered on time and the situations encountered during the delivery process, will be recorded and analyzed by the platform. The platform uses machine learning technology to conduct in-depth analysis of all feedback data and delivery data, find out the deficiencies of the current evaluation model and rules, and regularly update and optimize the rules. The weight coefficients of factors affecting delivery time are continuously optimized through the gradient descent method. The expression formula is: S q =μ0+μ1G+μ2Mt +μ3Z+μ4J tfc , where S q is the predicted delivery time, G is the rider workload, M t is the weather condition, Z is the rider load, J tfc is the traffic condition, μ0, μ1, μ2, μ3, μ4 are coefficients, and the actual delivery time S sj and the predicted delivery time S q The difference is used to update the coefficient, and the expression formula is: Where α is the learning rate, is the gradient of the loss function. In order to combine the influence of historical data and the latest feedback data, the platform dynamically adjusts the prediction time through the weighted average method. The expression formula is: in is the predicted time based on historical data, is the prediction time based on feedback data, and E is the weight coefficient, which is used to balance the impact of historical data and real-time feedback. It is used to continuously adjust the hyperparameters of the model through Bayesian optimization. The expression formula is: Where f(θ) is the loss function of the model, and θ is a hyperparameter including but not limited to learning rate and regularization parameter.

[0076] In this embodiment, the platform continuously collects and analyzes the riders' work data, including but not limited to the route, time, difficulties encountered, and handling methods of each delivery. Based on the continuously accumulated data, the platform identifies the key factors affecting delivery efficiency, and then optimizes the riders' personalized work ability evaluation model through machine learning technology. Based on the data feedback after each delivery, the optimal delivery mode is obtained, thereby further improving the processing method of future orders, which is conducive to optimizing delivery strategies and route planning, ensuring that each delivery is completed in the shortest possible time, reducing the burden on riders, and improving the adaptability of the platform.

[0077] The above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit the same. Although the present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that the technical solutions described in the aforementioned embodiments may still be modified, or some of the technical features may be replaced by equivalents. Such modifications or replacements will not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. An optimization method formed by secondary feedback from riders, characterized in that: include: S1. Build an Internet-connected platform to collect various information of riders in real time; S2. Establish a feedback mechanism so that riders can provide real-time feedback on important information encountered during the delivery process; S3. Collect riders’ work data and generate a personalized work ability assessment model; S4. Networked map software collects road information and combines feedback data with historical data to establish a route prediction mechanism; S5. Real-time monitoring of rider delivery information and providing solutions in case of emergencies; S6. Obtain delivery information, learn and optimize the rider’s evaluation model and route planning.

2. The optimization method according to claim 1, characterized in that: In step S1, the method of building a network platform to collect various information of riders in real time is: Build an Internet-connected platform that connects food delivery platforms, riders and consumers, and then use the platform to obtain real-time data on rider delivery orders. The data includes but is not limited to order generation time, delivery distance, restaurant location and delivery address. At the same time, the Internet-connected platform will obtain the rider's daily workload and load information in real time during the delivery process, and then upload the rider's real-time location, driving speed and road condition information to the platform through IoT devices.

3. The optimization method formed by secondary feedback from the rider according to claim 2, characterized in that: In step S2, a feedback mechanism is established, and the method for the rider to provide real-time feedback on important information encountered during the delivery process is as follows: Riders use applications on their mobile internet-connected devices to provide feedback during the delivery process on issues including but not limited to traffic jams, road closures, long wait times for food, and delays in elevator use. Feedback can be provided by clicking on preset options or through voice input.

4. The optimization method formed by secondary feedback from the rider according to claim 3, characterized in that: In step S2, a feedback mechanism is established, and the method for the rider to provide real-time feedback on important information encountered during the delivery process is as follows: When a rider encounters an emergency, the feedback delay time is expressed as: S T (n) = x n ×F n , where S T (n) represents the delay time, F n represents the impact time of the nth feedback event, χ n It is a coefficient set according to the feedback type. When a rider has multiple feedbacks in the same delivery process, the cumulative feedback delay time is expressed as: Where m is the total number of events reported by the rider, S T is the accumulated feedback delay time.

5. The optimization method formed by secondary feedback from the rider according to claim 4, characterized in that: In step S3, the method of collecting the rider's work data and generating a personalized work ability evaluation model is as follows: The platform collects the riders' historical work performance, including but not limited to average delivery speed, ability to cope with different weather and traffic conditions, and work efficiency during single and multiple deliveries. The platform combines these historical data with current order data, including but not limited to delivery distance, weather, and traffic routes, and uses machine learning algorithms to generate personalized delivery time recommendations, creating a personalized work ability assessment model for each rider.

6. The optimization method formed by secondary feedback from the rider according to claim 5, characterized in that: In step S3, the method of collecting the rider's work data and generating a personalized work ability evaluation model is as follows: Taking into account the rider's historical data, the day's workload, and environmental factors, the formula for evaluating the rider's delivery time is: q =A base ×(1-ζ1×M-ζ2×S T -ζ3×I), where Α q is the delivery speed of the rider, A base is the basic riding speed of the rider, which is taken from his historical data. M is the cumulative workload of the rider on that day, including but not limited to the number of delivered orders. S T is the influencing factor of the current weather conditions, including but not limited to deceleration in rainy days, waiting for traffic lights and road congestion. I is the load of a single delivery, including but not limited to the weight of the order or the number of deliveries. ζ1, ζ2 and ζ3 are the corresponding influencing coefficients, which can be obtained through regression analysis of historical data. The delivery time of the rider will be affected by the speed and delivery distance. The expression formula of the rider's personalized delivery time is: Where Q L is the personalized delivery time, X is the delivery distance, A q is the rider’s personalized speed, G xh Other fixed time consumption includes but is not limited to parking, going up and down elevators, and handing over items. When multiple influencing factors need to be considered comprehensively, the formula for expressing the rider's personalized delivery time can be expanded to: Among them G jt It is the influencing factor of traffic conditions and is obtained through traffic data.

7. The optimization method according to claim 6, characterized in that: In step S4, the networked map software collects road information and combines feedback data and historical data to establish a route prediction mechanism: After receiving the feedback data, the platform immediately calculates the impact of these emergencies on the delivery time. The platform adjusts the estimated delivery time of the current order based on the feedback data, and notifies the user and the food delivery platform for secondary feedback when necessary. The platform can also optimize the traffic route selection based on repeated emergencies to avoid similar situations from happening again. The expression formula for the estimated delivery time combined with the feedback delay is: S fkh =S zcsj +S T , where S zcsj is the initial delivery time evaluation result without considering the feedback factor, S T is the accumulated delay introduced by rider feedback.

8. The optimization method according to claim 7 formed by secondary feedback from the rider, characterized in that: In step S4, the networked map software collects road information and combines feedback data and historical data to establish a route prediction mechanism: The platform's networked map software takes into account the current situation, including but not limited to traffic conditions, weather conditions, and road construction, and dynamically optimizes the rider's delivery route. When the rider reports in real time that there is traffic congestion or a blocked road ahead, the platform's networked map software immediately gives the best route again and recommends detours to the rider to reduce delivery delays. In addition, the platform can predict traffic conditions in specific time periods and locations by analyzing historical traffic data from the map software, and optimize the rider's delivery route in advance to avoid congestion during peak hours. The platform's optimized route expression formula is: E = ZXT d (B), where B is the Bth alternative route, T d (B) depends on the conditions of each route, including but not limited to traffic conditions, road length, and weather. During the delivery process, if there is an emergency in the road conditions, the platform will adjust the delivery route in real time based on the feedback. The expression formula is: S new =ZX[T d (B)+T d ], select route B with the least delay after adjustment, and use the historical traffic data to predict the traffic conditions in a certain period of time in the future: S tf (t) = ∫yzt(t, X, G, J), where ∫yzt is a traffic prediction function based on historical data modeling, t is the estimated time, X is the route distance, G is the weekday or weekend influencing factor, and J is the time period including but not limited to peak hours and non-peak hours. The platform uses a comprehensive feedback mechanism, personalized evaluation and route prediction to obtain the final dynamic optimization of delivery time expression formula: Among them A T (B) is the rider’s personalized speed on route t, T d (B) is the delay time of the route, including but not limited to the influence of traffic, weather, traffic lights and merchant’s food delivery feedback.

9. The optimization method formed by secondary feedback from the rider according to claim 8, characterized in that: In step S5, the method of real-time monitoring of the rider's delivery information and providing a solution in the face of emergencies is as follows: Based on the data from the riders’ real-time feedback, including but not limited to traffic jams and bad weather, the platform automatically sends delay notification information to the food delivery platform and users when the estimated delivery time is not within the predetermined standard. Users can view the latest estimated delivery time through the platform and choose to continue waiting or cancel the order based on the delay. The platform can automatically split the order based on the riders’ feedback and assign part of the delivery tasks to other nearby riders to reduce the workload of the current riders. At the same time, the platform adjusts the delivery order of multiple orders based on the riders’ secondary feedback data, giving priority to orders that are closer or with tight estimated time. When the platform finds a delay, the expression formula for determining which orders need to be prioritized is: Where Z p is the order priority, D dly is the delay time of the order, X is the delivery distance of the order, J p The order value includes but is not limited to orders with larger amounts, which have higher priority. κ1, κ2, and κ3 are weight coefficients. m is the current number of orders for the rider. When the platform finds that the rider cannot complete the task, the expression formula for reallocation is: Among them I e is the current rider's load, X e is the distance between the rider and the order destination, τ1, τ2 are weight coefficients, y is the total number of currently available riders, and the platform automatically updates the estimated delivery time as follows: U x =U yj +U dly +U aut , where U x is the estimated delivery time after automatic update. yj is the current estimated delivery time, U dly is the delay time estimated based on real-time feedback, U aut It is the additional time after the system reallocates or adjusts the route. The platform splits an order into multiple parts according to the expression formula: Among them, H i is the cost function of order splitting, S q (n) is the delivery time of the order part, and q is the number of parts after the order is split.

10. The optimization method formed by secondary feedback from the rider according to claim 9, characterized in that: In step S6, the method of real-time monitoring of the rider's delivery information and providing a solution in the face of emergencies is as follows: Every feedback from the riders and the delivery results of each order, including but not limited to whether it is delivered on time and the situations encountered during the delivery process, will be recorded and analyzed by the platform. The platform uses machine learning technology to conduct in-depth analysis of all feedback data and delivery data, identify the deficiencies of the current evaluation model and rules, and regularly update and optimize the rules. The weight coefficients of factors affecting delivery time are continuously optimized through the gradient descent method. The expression formula is: S q =μ0+μ1G+μ2M t +μ3Z+μ4J tfc , where S q is the predicted delivery time, G is the rider workload, M t is the weather condition, Z is the rider load, J tfc is the traffic condition, μ0, μ1, μ2, μ3, μ4 are coefficients, and the actual delivery time S sj and the predicted delivery time S q The difference is used to update the coefficients, and the expression formula is: Where α is the learning rate, is the gradient of the loss function. In order to combine the influence of historical data and the latest feedback data, the platform dynamically adjusts the prediction time through the weighted average method. The expression formula is: in is the predicted time based on historical data, is the prediction time based on feedback data, l is the weight coefficient, which is used to balance the impact of historical data and real-time feedback. It is used to continuously adjust the hyperparameters of the model through Bayesian optimization. The expression formula is: in is the loss function of the model, Hyperparameters include but are not limited to learning rate and regularization parameters.

Images