Dispatch Management System
The dispatch management system addresses inconsistent service quality by integrating real-time traffic and weather data into fare calculations, optimizing driver selection, and enhancing transportation monitoring, resulting in improved dispatch efficiency and service quality.
Patent Information
- Application Number
- KR1020250092194
- Authority / Receiving Office
- KR · KR
- Patent Type
- Patents
- Current Assignee / Owner
- Priority Date
- 2025-07-01
- Filing Date
- 2025-07-09
- Publication Date
- 2026-11-06
- Estimated Expiration
- 2045-07-09
AI Technical Summary
Existing dispatch systems fail to effectively integrate real-time traffic, weather, and time of day variables into fare calculations, leading to inconsistent service quality and limited management capabilities, and lack real-time transportation status monitoring and feedback incorporation.
A dispatch management system that includes a management server for real-time fare calculation, driver recommendation, transportation management, evaluation, and report generation, utilizing multivariate data and external environment factors to optimize dispatch processes.
Enables flexible, real-time fare calculation and intelligent dispatch management, improving service quality through dynamic pricing, driver selection, and comprehensive transportation monitoring.
Smart Images

Figure 112025077406656-PAT00001_ABST
Abstract
Description
Technology Field
[0001] The present invention relates to a dispatch management system. Background Technology
[0002] As the demand for logistics and transportation services increases rapidly in modern society, the efficiency and speed of dispatch management are emerging as key factors determining the overall competitiveness of the logistics industry.
[0003] However, existing dispatch systems are based on a fixed fare structure and have failed to effectively reflect external variables such as real-time traffic conditions, weather, and time of day in fare calculations. Furthermore, because they fail to comprehensively consider various factors such as driver performance, ratings, and response speed during the dispatch process, there are issues of decreased customer satisfaction and variations in service quality.
[0004] Furthermore, the lack of features to monitor transportation status in real-time after dispatch or to efficiently incorporate feedback between drivers and customers into the system limits operators and managers' ability to comprehensively manage the overall transportation flow and identify areas for improvement.
[0005] Meanwhile, the aforementioned background technology is technical information that the inventor possessed for the derivation of the present invention or acquired during the process of deriving the present invention, and it cannot necessarily be considered publicly known technology disclosed to the general public prior to the filing of the present invention. Prior art literature
[0006] Korean Registered Patent No. 100653661 The problem to be solved
[0007] The objective of the present invention is to provide a dispatch management system capable of intelligently integrating and managing the entire process from dispatch to the completion of transportation.
[0008] The technical problems of the present invention are not limited to those mentioned above, and other unmentioned technical problems will be clearly understood by those skilled in the art from the description below. means of solving the problem
[0009] A dispatch management system according to one embodiment of the present invention may include a management server that calculates and provides real-time fares.
[0010] According to one embodiment, the management server may include: a fare calculation unit that extracts distance information and external environment data from dispatch request information received from a client terminal and automatically calculates dynamic fare information reflecting variable factors in the base fare by linking the extracted distance information and external environment data to a fare table; a dispatch recommendation unit that generates a driver recommendation list based on driver multivariate data including distance, rating, response time, and driving history among a list of previously registered drivers and provides dispatch information to the corresponding user terminal; a transportation management unit that provides real-time dispatch information in stages of reception, dispatch completion, driving, and completion, and collects location information of the user terminal in operation to generate visualization data through a map-based visualization UI and provides it to the administrator terminal and the client terminal in real time; an evaluation management unit that receives ratings and reviews for the driver who drove the corresponding transport from the client terminal where the transport has been completed and reflects them in the driver's driver multivariate data; and a report generation unit that automatically collects and integrates dispatch records, driving time, and fare information to generate settlement details and supports conversion to Excel format.
[0011] According to one embodiment, the user terminal refers to a terminal possessed by a registered driver that can receive dispatch information through the dispatch recommendation unit and accept or reject it, and when the dispatch information provided by the dispatch recommendation unit is accepted, real-time location information can be provided to the transportation management unit.
[0012] According to one embodiment, the administrator terminal may refer to a terminal possessed by an administrator who manages or operates the management server, and capable of displaying the visualization data generated by the transportation management unit.
[0013] According to one embodiment, the client terminal may refer to a terminal possessed by a client who has requested the transportation of goods through the management server, and capable of displaying the visualization data generated by the transportation management unit.
[0014] According to one embodiment, the distance information may refer to the distance information between delivery points from the origin information and the destination information.
[0015] According to one embodiment, the external environment data may refer to external factors affected during dispatch operation, including traffic congestion, weather, time of day, and road conditions collected in real time through an API that provides information on traffic congestion, weather, time of day, and road conditions.
[0016] According to one embodiment, the fare table may refer to a reference database or matrix for calculating the final fare by reflecting various variable factors (e.g., distance, traffic, weather) based on a base fare.
[0017] According to one embodiment, the freight calculation unit may, when there are multiple (N) destinations, set the destination closest to the origin information as the first destination and the destination farthest from the origin information as the Nth destination, and collect delivery distance information between the origin information and the first destination information, delivery distance information between the first destination and the second destination information, and delivery distance information between the second destination and the third destination information.
[0018] According to one embodiment, the dispatch recommendation unit can generate the article recommendation list by scoring the article multivariate data and arranging the article multivariate data with a high score to be placed at the top for priority recommendation.
[0019] According to one embodiment, the dispatch recommendation unit can provide dispatch information to the user terminal, and if the user terminal selects to accept the dispatch information, the dispatch information can be switched to a dispatch completion state in real time.
[0020] According to one embodiment, the transportation management unit can provide dispatch issues including intermediate transportation deviations and delays through the stages of reception, dispatch completion, operation, and completion, which are the real-time dispatch information.
[0021] According to one embodiment, the evaluation management unit analyzes the cumulative rating and review keywords received from the client terminal to continuously track and manage the service quality so that it is maintained or improved, and can automatically issue a warning to articles with a rating below a certain standard that are repeated based on the scored article multivariate data, or reflect this in the article multivariate data so that such articles become eligible for exclusion from dispatch.
[0022] According to one embodiment, the freight calculation unit comprises a delivery distance (D fc ), traffic congestion index (T fc ), Weather Influence Index (W fc ), time zone(H fc ), distance weight (α fc ), external environment weight (β fc ), time-weighted ratio (γ fc A fare calculation algorithm can be generated that reflects distance and external environmental data (traffic, weather, time of day, road conditions) using ) as elements.
[0023] According to one embodiment, the fare calculation algorithm may be composed of a combination of three terms, and the combination of three terms may be characterized by being composed of the sum of the first and second terms and the product of the third term.
[0024] According to one embodiment, the freight calculation algorithm comprises a delivery distance (D fc Distance weight (α) to ) fc The first term, designed to represent the basic fare by multiplying by ), and the traffic congestion index (T fc ) and weather impact index (W fc After applying a logarithmic function to the value multiplied by ), the external environment weight (β fc A second term designed to normalize the influence of the external environment by multiplying by ), and the time zone (H fc After adjusting by applying ) to the trigonometric function sine, the time weight ratio (γ fc It can be characterized by being composed of a combination of third terms designed to reflect changes in weighting during late-night or congested hours, with the value obtained by multiplying by ) and adding 1.
[0025] According to one embodiment, the fare calculation algorithm may include a second term characterized by using a logarithmic function to express a gradual increase in effect, as abrupt changes in traffic and weather have a non-linear effect.
[0026] An integrated dispatch and settlement automation system according to another embodiment of the present invention may include a management server that calculates and provides real-time fares based on distance information between a departure point and a destination and external environment data.
[0027] According to one embodiment, the management server may include: a fare calculation unit that extracts distance information and external environment data from dispatch request information received from a client terminal and automatically calculates dynamic fare information reflecting variable factors in the base fare by linking the extracted distance information and external environment data to a fare table; a dispatch recommendation unit that generates a driver recommendation list based on driver multivariate data including distance, rating, response time, and driving history among a list of previously registered drivers and provides dispatch information to the corresponding user terminal; a transportation management unit that provides real-time dispatch information in stages of reception, dispatch completion, driving, and completion, and collects location information of the user terminal in operation to generate visualization data through a map-based visualization UI and provides it to the administrator terminal and the client terminal in real time; an evaluation management unit that receives ratings and reviews for the driver who drove the corresponding transport from the client terminal where the transport has been completed and reflects them in the driver's driver multivariate data; and a report generation unit that automatically collects and integrates dispatch records, driving time, and fare information to generate settlement details and supports conversion to Excel format.
[0028] According to one embodiment, the user terminal refers to a terminal possessed by a registered driver that can receive dispatch information through the dispatch recommendation unit and accept or reject it, and when the dispatch information provided by the dispatch recommendation unit is accepted, real-time location information can be provided to the transportation management unit.
[0029] According to one embodiment, the administrator terminal may refer to a terminal possessed by an administrator who manages or operates the management server, and capable of displaying the visualization data generated by the transportation management unit.
[0030] According to one embodiment, the client terminal may refer to a terminal possessed by a client who has requested the transportation of goods through the management server, and capable of displaying the visualization data generated by the transportation management unit.
[0031] According to one embodiment, the distance information may refer to the distance information between delivery points from the origin information and the destination information.
[0032] According to one embodiment, the external environment data may refer to external factors affected during dispatch operation, including traffic congestion, weather, time of day, and road conditions collected in real time through an API that provides information on traffic congestion, weather, time of day, and road conditions.
[0033] According to one embodiment, the fare table may refer to a reference database or matrix for calculating the final fare by reflecting various variable factors (e.g., distance, traffic, weather) based on a base fare.
[0034] According to one embodiment, the freight calculation unit may, when there are multiple (N) destinations, set the destination closest to the origin information as the first destination and the destination farthest from the origin information as the Nth destination, and collect delivery distance information between the origin information and the first destination information, delivery distance information between the first destination and the second destination information, and delivery distance information between the second destination and the third destination information.
[0035] According to one embodiment, the dispatch recommendation unit can generate the article recommendation list by scoring the article multivariate data and arranging the article multivariate data with a high score to be placed at the top for priority recommendation.
[0036] According to one embodiment, the dispatch recommendation unit can provide dispatch information to the user terminal, and if the user terminal selects to accept the dispatch information, the dispatch information can be switched to a dispatch completion state in real time.
[0037] According to one embodiment, the transportation management unit can provide dispatch issues including intermediate transportation deviations and delays through the stages of reception, dispatch completion, operation, and completion, which are the real-time dispatch information.
[0038] According to one embodiment, the evaluation management unit analyzes the cumulative rating and review keywords received from the client terminal to continuously track and manage the service quality so that it is maintained or improved, and can automatically issue a warning to articles with a rating below a certain standard that are repeated based on the scored article multivariate data, or reflect this in the article multivariate data so that such articles become eligible for exclusion from dispatch.
[0039] According to one embodiment, the freight calculation unit comprises a delivery distance (D fc ), traffic congestion index (T fc ), Weather Influence Index (W fc ), time zone(H fc ), distance weight (α fc ), external environment weight (β fc ), time-weighted ratio (γ fc A fare calculation algorithm can be generated that reflects distance and external environmental data (traffic, weather, time of day, road conditions) using ) as elements.
[0040] According to one embodiment, the fare calculation algorithm may be composed of a combination of three terms, and the combination of three terms may be characterized by being composed of the sum of the first and second terms and the product of the third term.
[0041] According to one embodiment, the freight calculation algorithm comprises a delivery distance (D fc Distance weight (α) to ) fcThe first term, designed to represent the basic fare by multiplying by ), and the traffic congestion index (T fc ) and weather impact index (W fc After applying a logarithmic function to the value multiplied by ), the external environment weight (β fc A second term designed to normalize the influence of the external environment by multiplying by ), and the time zone (H fc After adjusting by applying ) to the trigonometric function sine, the time weight ratio (γ fc It can be characterized by being composed of a combination of third terms designed to reflect changes in weighting during late-night or congested hours, with the value obtained by multiplying by ) and adding 1.
[0042] According to one embodiment, the fare calculation algorithm may include a second term characterized by using a logarithmic function to express a gradual increase in effect, as abrupt changes in traffic and weather have a non-linear effect. Effects of the invention
[0043] According to one aspect of the present invention described above, the dispatch management system proposed by the present invention can overcome the disadvantages of a fixed fare system and enable flexible fare calculation that reflects real-time conditions of supply and demand.
[0044] In addition, it enables intelligent integrated management of the entire process from dispatch to transportation completion by automatically calculating fares based on real-time distance information and external environmental data, providing sophisticated dispatch recommendations utilizing multivariate data on drivers, visualizing real-time transportation status, managing drivers based on ratings, and automatically settling payments and generating reports.
[0045] The effects of the present invention are not limited to those mentioned above, and various effects may be included within the scope obvious to a person skilled in the art from the contents described below. Brief explanation of the drawing
[0046] FIG. 1 is a conceptual diagram of a dispatch management system according to one embodiment of the present invention. FIG. 2 is a conceptual diagram of a management server according to one embodiment of the present invention. Specific details for implementing the invention
[0047] The following detailed description of the invention refers to the accompanying drawings, which illustrate specific embodiments in which the invention may be practiced. These embodiments are described in sufficient detail to enable those skilled in the art to practice the invention. It should be understood that various embodiments of the invention are different but need not be mutually exclusive. For example, specific shapes, structures, and characteristics described herein may be implemented in other embodiments without departing from the spirit and scope of the invention in relation to one embodiment.
[0048] When it is stated that one component is "connected" or "contracted" to another component, it should be understood that while it may be directly connected or contracted to that other component, there may also be other components in between. Conversely, when it is stated that one component is "directly connected" or "directly contracted" to another component, it should be understood that there are no other components in between.
[0049] Furthermore, it should be understood that the location or arrangement of individual components within each disclosed embodiment may be changed without departing from the spirit and scope of the invention. Accordingly, the following detailed description is not intended to be taken in a limiting sense, and the scope of the invention is limited only by the appended claims, including all equivalents thereof, provided appropriately described. Similar reference numerals in the drawings refer to the same or similar functions across various aspects.
[0050] Hereinafter, preferred embodiments of the present invention will be described in more detail with reference to the drawings.
[0052] FIG. 1 is a conceptual diagram of a dispatch management system according to one embodiment of the present invention.
[0053] Referring to FIG. 1, a dispatch management system according to one embodiment of the present invention may include a management server (100), a user terminal (300), an administrator terminal (500), and a client terminal (700).
[0054] The management server (100) is a core component that integrates and intelligently supports the entire dispatch and transportation process according to the present invention, and can perform roles such as calculating and providing real-time freight rates through various function modules.
[0055] The user terminal (300) can receive the dispatch information through the management server (100) and accept or reject it, and refers to a terminal possessed by a registered driver, and if the user accepts the dispatch information provided by the management server (100), it can provide real-time location information to the management server (100).
[0056] The administrator terminal (500) may refer to a terminal possessed by an administrator who manages or operates the management server (100) and capable of displaying the visualization data generated by the management server (100).
[0057] The client terminal (700) may refer to a terminal possessed by a client who has requested the transportation of goods through the management server (100), and capable of displaying the visualization data generated by the management server (100).
[0058] The management server (100), user terminal (300), administrator terminal (500), and client terminal (700) may be their own servers or cloud servers for providing the service according to the present invention, or they may be a peer-to-peer (P2P) set of distributed nodes.
[0059] The management server (100) can perform one or more of the operations, storage, reference, input / output, and control functions of a general computer, and may include an artificial neural network described later based on input data.
[0060] The management server (100) may include a processor and memory. The processor may include devices capable of integrally controlling and intelligently supporting the entire dispatch and transportation process according to the present invention, calculating and providing real-time freight rates, and performing these functions. The processor may execute a program or control the management server (100). Program code executed by the processor may be stored in memory. The memory may store relevant information for performing a service according to the present invention or a program for implementing a method. The memory may be volatile memory or non-volatile memory.
[0061] The management server (100) can send data to an external device or receive data from an external device using a network.
[0062] The management server (100) can train an artificial neural network and can also use an artificial neural network that has been trained. The processor can train or execute an artificial neural network stored in memory, and the memory can store an artificial neural network that has been trained. The electronic device that trains the artificial neural network and the electronic device that uses it may be the same, but they may also be separate.
[0063] Artificial intelligence is a computer system that partially implements the functions of the human brain and is capable of learning, speculating, and making judgments on its own. As learning progresses, the probability of extracting the correct answer can increase. Artificial intelligence can be composed of learning and component technologies that utilize it. The learning aspect of AI is an algorithmic technology that classifies and learns features based on input data, while the component technologies may be techniques that utilize these learning algorithms to partially implement the functions of the human brain.
[0064] Artificial intelligence is a technology that facilitates the approach to problems where multiple probabilistic answers are possible, enabling it to logically and probabilistically infer optimal cycles, methods, and plans based on input data. AI inference techniques can include evaluating input data, optimization prediction, knowledge and probability-based reasoning, and preference-based planning.
[0065] Artificial neural networks are learning algorithms in the field of machine learning that programmatically implement the connections between neurons and synapses in the brain. By creating a neural network structure through programming and then training it, artificial neural networks can acquire desired functions. Although errors may exist, they can learn from massive datasets to produce appropriate output data from input data. They have the advantage of being able to obtain output data that has yielded statistically good results and are similar to human reasoning.
[0066] The management server (100) can infer individual characteristics and interests by analyzing consumers' online behavior data, social media activities, search history, etc., using an artificial intelligence algorithm built based on big data, and may include a number of pre-trained artificial neural networks for this purpose.
[0067] The network is a high-speed backbone network of a large-scale communication network capable of high-capacity, long-distance voice and data services, and may be a next-generation wired and wireless network for providing the Internet or high-speed multimedia services.
[0068] If the network is a mobile communication network, it may be a synchronous mobile communication network or an asynchronous mobile communication network. As an example of an asynchronous mobile communication network, a WCDMA (Wideband Code Division Multiple Access) network may be cited. In this case, although not shown in the drawing, the network may include an RNC (Radio Network Controller). Meanwhile, although a WCDMA network was given as an example, it may be a 3G LTE network, a 4G network, a next-generation communication network such as 5G, or other IP-based IP networks.
[0069] The management server (100), user terminal (300), administrator terminal (500), and client terminal (700) may include any terminal capable of exchanging data over a network, such as a desktop computer, laptop, tablet, or smartphone.
[0070] The management server (100), user terminal (300), administrator terminal (500), and client terminal (700) may include one or more of the computational function, storage function, reference function, input / output function, and control function of a computer to perform the service according to the present invention.
[0071] The management server (100), user terminal (300), administrator terminal (500), and client terminal (700) may access a website or install an application to receive the service according to the present invention. The management server (100), user terminal (300), administrator terminal (500), and client terminal (700) may exchange data through the website or the application.
[0072] The network is a high-speed backbone network of a large-scale communication network capable of high-capacity, long-distance voice and data services, and may be a next-generation wired and wireless network for providing the Internet or high-speed multimedia services.
[0073] If the network is a mobile communication network, it may be a synchronous mobile communication network or an asynchronous mobile communication network. As an example of an asynchronous mobile communication network, a WCDMA (Wideband Code Division Multiple Access) network may be cited. In this case, although not shown in the drawing, the network (300) may include an RNC (Radio Network Controller). Meanwhile, although a WCDMA network was given as an example, it may be a 3G LTE network, a 4G network, a 5G network, or other next-generation communication networks, or other IP-based IP networks.
[0074] The system (1) according to one embodiment of the present invention can play a key role in improving the overall efficiency and reliability of the logistics system by operating the entire process from receiving a dispatch request to calculating the fare, assigning a driver, tracking transportation, post-evaluation, and settlement in real-time, automated, and data-based manner.
[0076] FIG. 2 is a conceptual diagram of a management server according to one embodiment of the present invention.
[0077] Referring to FIG. 2, a management server (100) according to one embodiment of the present invention may include a fare calculation unit (110), a dispatch recommendation unit (130), a transportation management unit (150), an evaluation management unit (170), and a report generation unit (190).
[0078] The fare calculation unit (110) can extract distance information and external environment data from dispatch request information received from the client terminal (700), and automatically calculate dynamic fare information that reflects variable factors in the basic fare by linking the extracted distance information and external environment data to a fare table.
[0079] The aforementioned distance information may refer to the distance information between deliveries from the origin information and the destination information.
[0080] In addition, the aforementioned external environment data may refer to external factors affected during dispatch operations, including traffic congestion, weather, time of day, and road conditions collected in real time through an API that provides information on traffic congestion, weather, time of day, and road conditions.
[0081] In addition, the aforementioned fare table may refer to a reference database or matrix for calculating the final fare by reflecting various variable factors (e.g., distance, traffic, weather) based on a base fare.
[0082] When there are multiple (N) destinations, the freight calculation unit (110) can set the destination closest to the origin information as the first destination and the destination farthest from the origin information as the Nth destination, and collect delivery distance information between the origin information and the first destination information, delivery distance information between the first destination and the second destination information, and delivery distance information between the second destination and the third destination information.
[0083] Meanwhile, the fare calculation unit (110) can generate and utilize a fare calculation algorithm that reflects distance and external environment data (traffic, weather, time of day, road conditions) to calculate the fare.
[0084] The freight calculation algorithm is based on delivery distance (D fc ), traffic congestion index (T fc ), Weather Influence Index (W fc ), time zone(H fc ), distance weight (α fc ), external environment weight (β fc ), time-weighted ratio (γ fc It can refer to an algorithm designed to calculate freight charges using ) as an element.
[0085] Delivery distance (D fc) can mean the driving distance (km) between the starting point and the destination, and can be derived by calculating optimal route-based distance information that reflects real-time traffic from map APIs (Google Maps, Kakao, Tmap, etc.).
[0086] The freight calculation unit (110) is the delivery distance (D fc When deriving ), using the distance based on the actual route of transportation is more accurate than the simple straight-line distance, so it can be characterized by using a map API.
[0087] Traffic congestion index (T fc ) can represent an indicator of the degree of congestion in the current road conditions and can be derived as a value between 0 and 10, and the closer to 0, the smoother the road corresponding to the route between the starting point and the destination, and the closer to 10, the more severe the congestion.
[0088] The degree of congestion in the current road conditions described above is based on the traffic congestion index (T) when the user terminal (300) receives dispatch information, accepts the dispatch, and the dispatch information is converted to a dispatch completion state. fc It may mean deriving ).
[0089] In addition, the fare calculation unit (110) normalizes the speed per road provided by the real-time traffic API to obtain a traffic congestion index (T fc ) can be derived.
[0090] For example, if the real-time travel speed decreases by 30% compared to the average travel speed of the road corresponding to the route between the departure point and the arrival point that has been switched to a dispatched state, the fare calculation unit (110) [calculates] the traffic congestion index (T fc ) can be derived as 3.
[0091] Weather Impact Index (W fc) is a real value between 0 and 1, which can represent the coefficient of influence that weather factors, such as precipitation, visibility, and snowfall, have on transportation, and the closer to 0, the less the weather influences, and the closer to 1, the worse the weather.
[0092] The fare calculation unit (110) can collect real-time weather data from an external weather API (weather agency, OpenWeather, etc.) and, based on precipitation (rain, snow), visibility, temperature (risk increases when below freezing), and road surface conditions, a weather impact index (W fc ) can be derived.
[0093] For example, if it is raining and the temperature is below zero based on real-time weather data collected from an external weather API, the freight calculation unit (110) adds 0.1 from 0, which represents the best weather, to 0.2, thereby increasing the weather influence index (W fc ) can be derived.
[0094] Time zone (H fc ) can represent a fare adjustment factor based on increased demand by time of day, such as late-night hours or commuting hours, and can be derived by multiplying the current time by pi (π) and dividing by 12 to reflect a daily cycle (0~2π).
[0095] For example, at commuting times of 6:00 or 18:00, the time zone (H fc ) becomes a value of π / 2 or 3π / 2, and is applied to the trigonometric function sine so that the fare can be adjusted to 1.3 times or 0.7 times.
[0096] Distance weight (α fc ) does not use any units and may refer to factors determining the fare per unit distance, and may refer to values derived based on basic fare policies or historical data analysis.
[0097] For example, if set to 1,000 won per 1km, the distance weight (α fc ) can be derived as 1,000 won.
[0098] External environment weight (β fc There is no unit used, and it may represent a weight applied to additional costs due to traffic and weather data, and it can be derived by reverse-calculating the fare increase based on past operation cases (delays, rain, traffic congestion, etc.).
[0099] For example, traffic congestion index (T fc ) is 10, weather impact index (W fc ) is when 1
[0100] Used with the logarithmic function, the traffic congestion index (T fc ) and weather impact index (W fc It can moderately reflect the increase effect of ).
[0101] Time-weighted ratio (γ fc ) has no unit used and may refer to a coefficient that determines how sensitively fares change depending on the time of day, and may refer to a value derived by analyzing how much higher the average fare was during rush hour and night hours through an analysis of past demand / fare change rates by time of day.
[0102] As described above, each element used in the freight calculation algorithm can be quantified and applied from real-time collectible API data, historical transportation data, or historical freight data.
[0103] Each element used in the fare calculation algorithm can be dynamically set according to analysis or policy criteria.
[0104] In addition, each element used in the freight calculation algorithm reflects realistic factors affecting freight rates and can be characterized as being used to enhance the efficiency and fairness of transportation services.
[0105] The fare calculation algorithm can be composed of a combination of three terms, and the combination of three terms can be characterized by being composed of the sum of the first and second terms and the product of the third term.
[0106] The freight calculation algorithm is based on delivery distance (D fc Distance weight (α) to ) fc The first term, designed to represent the basic fare by multiplying by ), and the traffic congestion index (T fc ) and weather impact index (W fc After applying a logarithmic function to the value multiplied by ), the external environment weight (β fc A second term designed to normalize the influence of the external environment by multiplying by ), and the time zone (H fc After adjusting by applying ) to the trigonometric function sine, the time weight ratio (γ fc It can be characterized by being composed of a combination of third terms designed to reflect changes in weighting during late-night or congested hours, with the value obtained by multiplying by ) and adding 1.
[0107] The second term of the fare calculation algorithm may be characterized by using a logarithmic function to express the gradual increase in impact, as abrupt changes in traffic and weather have a non-linear effect.
[0108] In other words, the fare calculation algorithm utilizes the second term to reduce the sensitivity to traffic and weather outliers and changes them to a controllable level, thereby producing the resulting fare (F fc It can be characterized by what is reflected in ).
[0109] In addition, the third term of the fare calculation algorithm may be characterized by using the trigonometric function sine to smoothly and periodically reflect time zone changes over 24 hours a day.
[0110] In other words, the fare calculation algorithm can utilize the third term to automatically reflect rates during rush hour or late-night hours.
[0111] The example of the freight calculation algorithm described above can be explained by substituting numerical values as follows.
[0112] Delivery distance (D fc Based on a distance of 12km and high traffic congestion, the traffic congestion index (T) is a score between 0 and 10. fc ) is 7, the weather impact index (W) derived based on rainy days fc ) is based on 0.6 and 18:00, time zone (H fc ) are derived as 3π / 2 respectively, and the distance weight (α fc ) is 1000, external environment weight (β fc ) is 1500, time weight ratio (γ fc When ) is set to 0.15, the fare (F fc ) can be calculated as 12,301.
[0113] The freight calculation unit (110) calculates the freight (F) calculated by the freight calculation algorithm. fc The fare for the corresponding dispatch can be calculated as 12,301 won based on the value of 12,301 of ), and the calculated dynamic fare information can be provided to the user terminal (300).
[0114] The freight calculation unit (110) measures the distance between GPS-based origin information and destination information and the delivery distance (D fc It can derive ), and collect traffic and weather information from the API to derive the traffic congestion index (T fc ) and weather impact index (W fc ) can be derived, and based on server time, the time zone (H fc ) can be derived and substituted as an input value into the freight calculation algorithm.
[0115] The fare calculation algorithm applies the input value received from the fare calculation unit (110) to the designed logarithmic and trigonometric functions, and then applies a distance weight (α) to each term. fc ), external environment weight (β fc ), time-weighted ratio (γ fcWeighting ) to freight (F fc ) can be produced.
[0116] The freight calculation unit (110) calculates the freight (F) from the freight calculation algorithm. fc ) can be displayed on the UI and provided to the user terminal (300), administrator terminal (500), and client terminal (700).
[0117] The freight calculation unit (110) calculates the delivery distance (D) at a predetermined interval (e.g., when the dispatch is completed). fc ), traffic congestion index (T fc ) and weather impact index (W fc Can update ).
[0118] Accordingly, the fare calculation unit (110) is in the time period (H) when it is switched to a dispatch completion state. fc Freight (F) depending on ) fc ) can change, and the value of the trigonometric function sine changes, so it can be subtracted at night and added during the day.
[0119] The fare calculation algorithm is intended to dynamically calculate the fare by reflecting real-time external environment and distance information in the fare calculation unit (110). Unlike the existing fixed rate method, it can validate the fare deviation by quantifying and normalizing supply-demand conditions and external factors into quantitative variables.
[0120] This can provide users with fair and real-time rates and administrators with prediction-based revenue optimization.
[0121] When the fare calculation unit (110) uses a fare calculation algorithm, it can set a fare that is 20-30% higher than the existing simple distance-based fare in case of urban traffic congestion or bad weather, and can also reflect dynamic fare based on nighttime hours.
[0122] Descriptions and examples of the components and functions of the freight calculation algorithm are sufficiently provided to facilitate implementation, thereby enabling a person of ordinary skill to easily implement it.
[0123] The dispatch recommendation unit (130) can generate a list of recommended drivers based on driver multivariate data including distance, rating, response time, and driving history among the previously registered drivers, and provide dispatch information to the corresponding user terminal (300).
[0124] Additionally, the dispatch recommendation unit (130) can generate the article recommendation list by scoring the article multivariate data and arranging the articles with high scores to be placed at the top for priority recommendation.
[0125] The dispatch recommendation unit (130) can provide dispatch information to the user terminal (300), and if the user terminal (300) selects to accept the dispatch information, the dispatch information can be switched to a dispatch completion state in real time.
[0126] Here, switching the dispatch information to a dispatched state is intended to prevent duplication by having another driver accept the information, as the assignment has been completed for the driver who selected to accept the information.
[0127] Meanwhile, the dispatch recommendation unit (130) recommends a recommendation score (S) for each driver to improve the speed of dispatch and service quality through the selection of the optimal driver. i An article recommendation list generation algorithm capable of producing ) can be utilized.
[0128] The article recommendation list generation algorithm is the cumulative rating (R) of the article. i ), distance between current departure point information and the corresponding article (D i ), average response time of the article (T i ), maximum distance among all articles (D max ), maximum response time among all articles (Tmax Recommendation score per article (S) using the importance weight of rating (w1), importance weight of distance (w2), and importance weight of response time (w3) as elements i It can be characterized by being configured to produce ).
[0129] Cumulative rating of the article (R i There is no separate unit used, and it may mean the average value of the rating received from the client terminal (700) after the driver has completed the delivery, and the client terminal (700) may input the driver's rating as a real number between 1 and 5 points.
[0130] That is, the dispatch recommendation unit (130) aggregates evaluation data received from the client terminal (700) in the evaluation management unit (170) and the cumulative rating (R) of the article. i ) can be derived.
[0131] Distance between current departure point information and the corresponding article (D i ) can use the unit 'km' and may represent the actual distance between the starting point and the current location of the driver from the dispatch request information received from the client terminal (700).
[0132] More specifically, the dispatch recommendation unit (130) collects the location of each driver in real time through the GPS of the user terminal (300), then collects the departure information of the dispatch request information received through the client terminal (700), calculates the distance between the drivers' locations and the departure information, and calculates the distance (D) between the current departure information and the driver. i ) can be derived.
[0133] Average response time of the article (T i ) can use the unit of 'seconds' and may refer to the average time it takes for the driver to actually accept or reject a dispatch request when receiving it.
[0134] More specifically, the dispatch recommendation unit (130) calculates the difference between the time a dispatch request is provided to the user terminal (300) and the time a response or rejection is received from the user terminal (300), and calculates the average response time of each driver for each dispatch request as an average value, thereby calculating the average response time (T) of the driver. i ) can be derived.
[0135] Maximum distance from current starting point (D) out of all articles max ) may mean the furthest distance among the recommended drivers at the time of the current dispatch request.
[0136] Maximum response time among all articles (T max ) may refer to the response time of the driver with the longest average response time among the recommended drivers at the time of the current dispatch request.
[0137] The importance weight of the rating (w1), the importance weight of the distance (w2), and the importance weight of the response time (w3) may represent values that adjust the relative influence of each item on the total score.
[0138] More specifically, the dispatch recommendation unit (130) can receive each value directly from an administrator or operator managing the server as a manually adjusted value by the administrator setting for the importance weight of the rating (w1), the importance weight of the distance (w2), and the importance weight of the response time (w3), and apply them to the article recommendation list generation algorithm.
[0139] An administrator or operator managing the server can calculate and input the values of each weight so that the sum of the importance weight of the rating (w1), the importance weight of the distance (w2), and the importance weight of the response time (w3) becomes 1.0.
[0140] The article recommendation list generation algorithm may be characterized by being composed of the average of three terms.
[0141] The first term of the article recommendation list generation algorithm is the cumulative rating (R) of the article. i It can be characterized by being designed to smooth the deviation between scores using the log-normalized value of ).
[0142] The article recommendation list generation algorithm is the cumulative rating (R) of the article. i It can be characterized by using a logarithmic function to reduce the range of change as the rating becomes higher, and through this, the recommendation score per article (S i It can prevent excessive influence on ).
[0143] The second term of the article recommendation list generation algorithm is the distance (D) between the current source information and the corresponding article. i It can be characterized by being designed to be inversely proportional using a cosine function so that the value increases as the article is closest to the source information based on the information.
[0144] More specifically, the second term is the distance between the current origin information and the corresponding article (D i ) the maximum distance from the current starting point among all articles (D max It can be divided by ) and multiplied by π / 2 and substituted into the trigonometric function cosine.
[0145] The article recommendation list generation algorithm may be characterized by being designed to efficiently reflect distance by using the trigonometric function cosine, such that articles closer to the source are closer to 1, and articles further away from the source decrease rapidly.
[0146] The third term of the article recommendation list generation algorithm is the article response time (T i It can be characterized by being designed with a normalized inverse term to give a higher score as the length is shorter.
[0147] More specifically, the third term is the average response time (T) of the article. i ) is the maximum response time (T) among all articlesmax It can be characterized by being designed as a simple linear reduction configuration that can reflect simple and intuitive time by applying the value obtained by dividing by ) and subtracting the value from 1.
[0148] Each term of the article recommendation list generation algorithm is adjusted by weights (w1, w2, w3) to obtain a final recommendation score (S i A single score called ) can be calculated, and the dispatch recommendation unit (130) can produce a recommendation score (S i You can sort articles and generate a recommendation list based on ).
[0149] Each weight (w1, w2, w3) can be designed to be flexibly adjustable, and in the case of urgent delivery, it can be characterized by being adjustable such that the distance weight (w2) is increased.
[0150] The example of the aforementioned article recommendation list generation algorithm, when converted to numerical values, is explained as follows.
[0151] Assuming there are articles A, B, and C, the cumulative rating of A (R i ) is 4.8, current distance between starting points (D i ) is 3, average response time (T i ) is a cumulative rating of 12, B (R i ) is 4.2, current distance between starting points (D i ) is 5, average response time (T i ) is the cumulative rating of 8, C (R i ) is 3.5, current distance between starting points (D i ) is 1, average response time (T i When ) is 20, the maximum distance from the current starting point among all articles (D max ) is 5, maximum response time among all articles (T max ) can be 20.
[0152] When the importance weight of rating (w1) is set to 0.4, the importance weight of distance (w2) is set to 0.35, and the importance weight of response time (w3) is set to 0.25, the recommendation score (S) of Article A i ) is 0.706, Article B's recommendation score (S i ) is 0.52, Article C's recommendation score (S i ) can be calculated as 0.674.
[0153] Calculated recommendation score per article (S i Based on ), the dispatch recommendation unit (130) selects the highest recommendation score (S i List Article A with ) at the top, and the recommendation score with the lowest score (S i Article B, which has ), can be listed at the bottom.
[0154] As described above, when the dispatch recommendation unit (130) generates a driver recommendation list, the recommendation score (S i Place the article with the highest ) at the top, and the recommendation score (S i You can create a list so that the article with the lowest value is placed at the bottom.
[0155] In addition, the dispatch recommendation unit (130) can collect the rating, distance, and response time of each article in real time and input them as input values to the article recommendation list generation algorithm.
[0156] The dispatch recommendation unit (130) can utilize an article recommendation list generation algorithm to perform normalization of ratings using a logarithmic function, distance transformation using a trigonometric function cosine, and normalization of response time for all articles.
[0157] The dispatch recommendation unit (130) can normalize the variables of each element through an article recommendation list generation algorithm, and then apply and sum the importance weight of the rating (w1), the importance weight of the distance (w2), and the importance weight of the response time (w3).
[0158] The dispatch recommendation unit (130) is a recommendation score (S) for each article calculated from an article recommendation list generation algorithm. i You can create a list of recommended articles by sorting them in order of highest value.
[0159] The dispatch recommendation unit (130) receives a dispatch request from the client terminal (700) and the cumulative rating (R) of the corresponding article. i ), distance between current departure point information and the corresponding article (D i ), average response time of the article (T i It can be derived by collecting ).
[0160] Through this, the dispatch recommendation unit (130) updates the driver-specific recommendation score (S) whenever a dispatch request is input from the client terminal (700). i You can generate an article recommendation list based on ).
[0161] Additionally, the dispatch recommendation unit (130) can decrease the importance weight (w1) of the rating and increase the importance weight (w3) of the response time when there is a surge in delivery request inputs from the client terminal (700).
[0162] As described above, the explanation and examples of the article recommendation list generation algorithm can be easily implemented by a person of ordinary skill in the art according to the definition and derivation methods of each component and element.
[0163] The transportation management department (150) provides real-time dispatch information in stages of receiving, dispatching, driving, and completion, and can collect location information of the user terminal (300) in operation to generate visualization data through a map-based visualization UI and provide it in real-time to the administrator terminal (500) and the client terminal (700).
[0164] The transportation management department (150) can provide dispatch issues including intermediate transportation deviations and delays through the stages of receiving, dispatching, driving, and completion, which are the real-time dispatch information mentioned above.
[0165] The evaluation management department (170) can receive ratings and reviews for the driver who performed the transportation from the client terminal (700) that completed the transportation and reflect them in the driver's multivariate data.
[0166] The evaluation management department (170) analyzes the cumulative rating and review keywords received from the client terminal (700) to continuously track and manage the service quality so that it is maintained or improved, and can automatically issue a warning to articles with a rating below a certain standard based on the scored article multivariate data or reflect this in the article multivariate data so that such articles are excluded from dispatch.
[0167] Meanwhile, the evaluation management department (170) integrates and quantifies multivariate data such as the article's rating, review keywords, and driving history to determine whether to issue a warning or exclude it from dispatch, and provides a comprehensive evaluation index (S me A comprehensive evaluation algorithm can be utilized to generate ).
[0168] The comprehensive evaluation algorithm uses the cumulative average rating (R) for articles. me ), Number of positive review keywords (K me ), maximum number of review keywords (K max ), problem number of runs (D me ), weight(w x , w y , w z A comprehensive evaluation index (S) for comprehensively quantitatively evaluating articles using ) as a factor and determining whether to issue a warning or exclude them from dispatch. me It can be characterized by being configured to produce ).
[0169] Cumulative average rating for the article (R me ) can use 'point' units and may represent the cumulative average of ratings entered through a client terminal for a specific article, and is composed of a range from 1.0 point (lowest) to 5.0 point (highest), and can be used as the most direct indicator of service satisfaction.
[0170] More specifically, the evaluation management department (170) receives rating reviews from client terminals (700) that have completed the client's transportation, aggregates them, and divides the total number of cases by the total score to obtain a cumulative average rating (R) for the article. me ) can be derived.
[0171] Cumulative average rating (R me ) is transformed through a logarithmic function, designed to have a greater impact on the score when the rating is low and to reflect the increase in the score gradually when the rating is high, thereby preventing distortion caused by extreme fluctuations in the rating.
[0172] Number of positive review keywords (K me ) can use the unit of ‘piece’ and can mean the number of keywords containing positive sentiment in the review text entered through the client terminal (700), and can be derived by applying a natural language processing (NLP)-based sentiment analysis engine to the review text to extract and count words that match or are similar to positive keywords registered in the keyword dictionary.
[0173] For example, the number of positive review keywords (K me Words such as "kindness," "accuracy," "cleanliness," and "speed" may apply, and positive words can be initially set by the server administrator or operator.
[0174] More specifically, the evaluation management unit (170) receives review sentences written on client terminals in real time, automatically separates positive words through a built-in keyword analysis module, and then applies a natural language processing (NLP)-based sentiment analysis engine to extract and count words that match or are similar to positive keywords registered in a keyword dictionary, thereby determining the number of positive review keywords (K me ) can be derived.
[0175] Maximum number of review keywords (K max) can use the unit of 'pieces' and can mean the maximum number of positive keywords recorded based on a single article among the registered articles, and can be derived through the maximum number of articles with the most collected positive keywords among the total article review records.
[0176] More specifically, the evaluation management department (170) analyzes the entire review database at a predetermined interval (e.g., once a day, every 100 reviews) and updates the maximum value to the maximum number of review keywords (K max ) can be derived.
[0177] Problem number of runs (D me ) may use the unit of 'pieces' and may refer to the number of runs performed by the driver that caused problems such as delays, deviations, or customer complaints.
[0178] The evaluation management department (170) determines whether there is abnormal operation based on GPS deviation, exceeding scheduled arrival time, and dispatch dissatisfaction feedback from clients in the operation records collected in real time by the transportation management department (150), and counts the number of problematic operations (D me ) can be derived.
[0179] weight(w x , w y , w z ) may refer to a value that quantifies the relative importance of each element (rating, review keywords, problem operation) in the score calculation, and may refer to a value that is directly set by the administrator or automatically adjusted and derived through the system's learning function.
[0180] The comprehensive evaluation algorithm can be composed of the sum of three terms.
[0181] The first term of the comprehensive evaluation algorithm is the cumulative average rating (R) for the article. me Based on the value reflecting ) in a logarithmic function, articles with higher ratings have a higher overall evaluation index (S meIt can be characterized by being designed to produce a high output.
[0182] More specifically, the comprehensive evaluation algorithm uses the cumulative average rating (R) for articles. me Since the effect of a change in ) is not the same when changing from 3 points to 4 points or from 4 points to 5 points, the effect of increasing the score can be adjusted smoothly by using a logarithmic function.
[0183] The second term of the comprehensive evaluation algorithm is the number of positive review keywords (K me ) maximum number of review keywords (K max It can be characterized by being designed to reflect an emotionally positive tendency by applying the trigonometric function sine after normalizing by the ratio.
[0184] More specifically, the comprehensive evaluation algorithm may be characterized by being designed so that the score naturally increases by applying a trigonometric curved change rather than a simple linear change to the review, which is an element of the client's sincerity and emotion.
[0185] The third term of the comprehensive evaluation algorithm is the number of problematic runs that occurred during operation (D me It can be characterized by being designed to reflect ) as a penalty by normalizing it with a logarithmic function.
[0186] More specifically, the comprehensive evaluation algorithm is the number of problem runs (D me Since the larger the value, the more it becomes a deduction factor, it can be characterized by being designed to provide a control function to prevent outliers (surges) and mitigate deductions through a logarithmic function.
[0187] Finally, the comprehensive evaluation algorithm may be characterized by reflecting weights to each element; since the importance of each element varies depending on the system or manager, this is referred to as a weight (w x , w y , w z It can be characterized by being designed to be flexibly adjustable by setting it as ).
[0188] The comprehensive evaluation algorithm can be used as a core component when the evaluation management department (170) analyzes article evaluation data, and can be utilized for automatic warning or dispatch exclusion processing functions for articles that repeatedly receive scores below a certain standard.
[0189] Although the existing simple average rating method had the potential for distortion, the evaluation management department (170) can improve fairness and precision by incorporating multivariate factors through the use of a comprehensive evaluation algorithm, and can achieve the effect of reducing the false warning rate by 22%.
[0190] The example of the comprehensive evaluation algorithm described above can be explained by substituting numerical values as follows.
[0191] Cumulative average rating for a specific article (R me ) is 4.2, number of positive review keywords (K me ) is 12, maximum number of review keywords (K max ) is 20, problem number of runs (D me ) is 5, weight(w x , w y , w z When ) is 1.0, 2.0, and 1.5, respectively, the comprehensive evaluation index (S me ) can be calculated as 0.86.
[0192] Comprehensive evaluation index (S) calculated from the above-mentioned numerical examples me ) can determine whether the article is subject to a warning or exclusion from dispatch based on a certain threshold value, and if the threshold value is 1.0, the article may be processed with one accumulated warning.
[0193] The aforementioned fixed threshold value can be received and stored as an initial setting value initially configured by the server administrator or operator.
[0194] In addition, the above-mentioned exclusion from dispatch may be determined by specific criteria (e.g., accumulation of 3 or more warnings), and the specific driver in the above-mentioned example may be automatically set to be excluded from dispatch when the total number of warnings reaches 3 as a result of processing 1 warning.
[0195] The evaluation management department (170) collects client reviews and rating inputs in real time and the cumulative average rating (R) for the article me It can derive ), and through the NLP review analyzer, the number of positive review keywords (K me ), maximum number of review keywords (K max ) can be derived, and the number of problem runs (D) from the transportation management department (150) me It can be derived by collecting ).
[0196] Afterwards, the evaluation management department (170) inputs each element derived from the comprehensive evaluation algorithm as an input value, applies it to a logarithmic function and a sine function, assigns weights, and performs calculations to obtain a comprehensive evaluation index (S me ) can be produced.
[0197] The evaluation management department (170) calculates the comprehensive evaluation index (S) based on the comprehensive evaluation algorithm. me It is possible to determine whether to issue a warning or exclude the article from dispatch by comparing it with a schedule threshold entered by the server administrator or operator.
[0198] The evaluation management department (170) uses a comprehensive evaluation algorithm at a predetermined interval (e.g., once a day after operation is completed) to evaluate each driver using a comprehensive evaluation index (S me ) can be produced.
[0199] In addition, the evaluation management department (170) has a comprehensive evaluation index (S me You can also analyze the trend of ratings for each article by tracking the average score of the past week along with the trend of change of ).
[0200] As described above, the explanation and examples of the comprehensive evaluation algorithm can be easily implemented by a person skilled in the art according to the definition and derivation methods of each component and element.
[0201] The report generation unit (190) can automatically collect and integrate dispatch records, driving times, and fare information to generate settlement details and support conversion into Excel format.
[0203] The embodiments described above are for illustrative purposes only, and those skilled in the art will understand that the embodiments described above can be easily modified into other specific forms without altering the technical concept or essential features of the embodiments described above. Therefore, the embodiments described above should be understood as illustrative in all respects and not restrictive. For example, each component described as a single unit may be implemented in a distributed manner, and components described as distributed may likewise be implemented in a combined form.
[0205] The scope of protection sought through this specification is defined by the claims set forth below rather than by the detailed description, and should be interpreted to include all modifications or variations derived from the meaning and scope of the claims and the concept of equivalents. Explanation of the symbols
[0206] 100: Management Server
Claims
Claim 1 A dispatch management system comprising a management server that calculates and provides real-time fares; wherein the management server comprises: a fare calculation unit that extracts distance information and external environment data from dispatch request information received from a client terminal, and automatically calculates dynamic fare information reflecting variable factors in the base fare by linking the extracted distance information and external environment data to a fare table; a dispatch recommendation unit that generates a driver recommendation list based on driver multivariate data including distance, rating, response time, and driving history from a list of previously registered drivers, and provides dispatch information to the corresponding user terminal; a transportation management unit that provides real-time dispatch information in stages of reception, dispatch completion, driving, and completion, and collects location information of a user terminal in operation to generate visualization data through a map-based visualization UI and provides it in real-time to an administrator terminal and the client terminal; and an evaluation management unit that receives ratings and reviews for the driver who performed the transportation from a client terminal where transportation has been completed and reflects them in the driver's driver multivariate data. It includes a report generation unit that automatically collects and integrates dispatch records, driving times, and fare information to generate settlement details and supports conversion to Excel format; the user terminal refers to a terminal possessed by a registered driver that can receive dispatch information through the dispatch recommendation unit and accept or reject it; if the dispatch information provided by the dispatch recommendation unit is accepted, it provides real-time location information to the transportation management unit; the manager terminal refers to a terminal possessed by a manager who manages or operates the management server and capable of displaying the visualization data generated by the transportation management unit; the client terminal refers to a terminal possessed by a client who requested the transportation of goods through the management server and capable of displaying the visualization data generated by the transportation management unit; the distance information refers to the distance information between deliveries from the origin information and the destination information; and the external environment data includes traffic congestion, weather, time of day,It refers to external factors affected during dispatch operation, including traffic congestion, weather, time of day, and road conditions collected in real-time through an API providing road condition information; the above-mentioned fare table refers to a reference database or matrix for calculating the final fare by reflecting various variable factors (e.g., distance, traffic, weather) based on a base fare; the above-mentioned fare calculation unit, when there are multiple (N) destinations, sets the destination closest to the above-mentioned origin information as the first destination and the destination farthest from the above-mentioned origin information as the Nth destination, collects delivery distance information between the above-mentioned origin information and the first destination information, delivery distance information between the above-mentioned first destination and the second destination information, and delivery distance information between the above-mentioned second destination and the third destination information; the above-mentioned dispatch recommendation unit generates the above-mentioned driver recommendation list by scoring the above-mentioned driver multivariate data and arranging drivers with high-scored driver multivariate data to be placed at the top for priority recommendation; and the above-mentioned dispatch recommendation unit provides dispatch information to the above-mentioned user terminal, and then receives a selection input from the said user terminal to accept the said dispatch information In this case, the dispatch information is converted to a dispatch completed state in real time, and the transportation management unit provides dispatch issues including intermediate transportation deviations and delays through the real-time dispatch information stages of reception, dispatch completed, operation, and completion, and the evaluation management unit analyzes the cumulative rating and review keywords received from the client terminal to continuously track and manage service quality so that it is maintained or improved, and automatically issues a warning to drivers whose ratings repeatedly fall below a certain standard based on the scored driver multivariate data or reflects this in the driver multivariate data to make them eligible for dispatch exclusion, and the freight calculation unit, delivery distance (D, fc ), traffic congestion index (T fc ), Weather Influence Index (W fc ), time zone(H fc ), distance weight (α fc ), external environment weight (β fc ), time-weighted ratio (γ fc A freight calculation algorithm is generated that reflects distance and external environmental data (traffic, weather, time of day, road conditions) using ) as an element, and the freight calculation algorithm can be composed of a combination of three terms, and the combination of three terms is characterized by being composed of the sum of the first and second terms and the product of the third term, and the freight calculation algorithm is characterized by delivery distance (D fc Distance weight (α) to ) fc The first term, designed to represent the basic fare by multiplying by ), and the traffic congestion index (T fc ) and weather impact index (W fc After applying a logarithmic function to the value multiplied by ), the external environment weight (β fc A second term designed to normalize the influence of the external environment by multiplying by ), and the time zone (H fc After adjusting by applying ) to the trigonometric function sine, the time weight ratio (γ fc It is characterized by being composed of a combination of a third term designed to reflect changes in weighting during late-night or congested hours by multiplying by ) and adding 1, and the fare calculation algorithm includes a second term characterized by using a logarithmic function to express a gradual increase in influence because rapid changes in traffic and weather have a non-linear effect, and the dispatch recommendation unit includes a recommendation score (S) for each driver i Improving the speed of dispatch and service quality through the selection of the optimal article using an article recommendation list generation algorithm that calculates ), and the said article recommendation list generation algorithm comprises the cumulative rating (R) of the relevant article i ), distance between current departure point information and the corresponding article (D i ), average response time of the article (T i ), maximum distance among all articles (D max ), maximum response time among all articles (T max Recommendation score per article (S) using the importance weight of rating (w1), importance weight of distance (w2), and importance weight of response time (w3) as elements i Characterized by being configured to calculate ), and the cumulative rating (R) of the said article i ) refers to the average value of the ratings received from the client terminal after the driver has completed the transportation, the client terminal inputs the driver's rating as a real number between 1 and 5 points, and the dispatch recommendation unit aggregates the evaluation data received from the client terminal by the evaluation management unit to determine the driver's cumulative rating (R i ) derive, and the distance (D) between the above current starting point information and the corresponding article. i ) uses the unit 'km' and refers to the actual distance between the origin and the current location of the corresponding driver based on the dispatch request information received from the client terminal, and the dispatch recommendation unit collects the location of each driver in real time through the GPS of the user terminal, collects the origin information of the dispatch request information received through the client terminal, calculates the distance between the drivers' locations and the origin information, and calculates the distance (D) between the current origin information and the corresponding driver. i ) derive, and the average response time (T) of the above article i ) uses the unit of 'seconds' and refers to the average time taken for the relevant driver to actually respond to accept or reject a dispatch request when receiving it, and the dispatch recommendation unit calculates the difference between the time the dispatch request was provided to the user terminal and the time the acceptance or rejection response arrived from the user terminal, wherein for each dispatch request, the response time of each driver is calculated as an average value to determine the average response time (T) of the relevant driver. i ) derive, and among all the above articles, the maximum distance (D) from the current starting point max ) refers to the farthest distance among the recommended drivers at the time of the current dispatch request, and the maximum response time (T) among all the above drivers. max ) refers to the response time of the article with the longest average response time among the recommended articles at the time of the current dispatch request, and the importance weight of the rating (w1), the importance weight of distance (w2), and the importance weight of response time (w3) refer to values that adjust the relative influence of each item on the total score, and the dispatch recommendation unit receives each value directly input by an administrator or operator managing the server as a value manually adjusted by the administrator settings for the importance weight of the rating (w1), the importance weight of distance (w2), and the importance weight of response time (w3), and applies them to the article recommendation list generation algorithm, and the first term of the article recommendation list generation algorithm is the cumulative rating (R) of the corresponding article i It is characterized by being designed to smooth the deviation between scores using the log-normalized value of ), and the article recommendation list generation algorithm comprises the cumulative rating (R) of the relevant article. i The above article-specific recommendation score (S) is characterized by using a logarithmic function to reduce the range of change as the rating becomes higher. i To prevent excessive influence on ), the second term of the above article recommendation list generation algorithm is the distance (D) between the current source information and the corresponding article. i It is characterized by being designed to be inversely proportional using a cosine function so that the value increases the closer the driver is to the origin information based on the ) information, wherein the distance (D) between the current origin information and the corresponding driver i ) the maximum distance from the current starting point among all articles (D max After dividing by ) and multiplying by π / 2, the result is substituted into the trigonometric function cosine, and the above article recommendation list generation algorithm is characterized by being designed to efficiently reflect distance by using the trigonometric function cosine such that articles closer to the source are closer to 1, and articles farther from the source decrease sharply, and the third term of the above article recommendation list generation algorithm is the article response time (T i It is characterized by being designed with a normalized inverse term to give a higher score as ) is shorter, wherein the average response time (T of the article) i ) is the maximum response time (T) among all articles max It is characterized by being designed with a simple linear reduction configuration to reflect simple and intuitive time by applying a value obtained by subtracting 1 from the value divided by ), and the article recommendation list generation algorithm adjusts the first to third terms with the weights (w1, w2, w3) to finally obtain a recommendation score (S i A single score is calculated as ), and the dispatch recommendation unit calculates the recommendation score (S) calculated from the article recommendation list generation algorithm. i Articles are sorted and a recommendation list is generated based on ), and each of the above weights (w1, w2, w3) is designed to be flexibly adjustable, characterized in that in the case of urgent delivery, the distance importance weight (w2) can be increased, and the dispatch recommendation unit, when generating the article recommendation list, the recommendation score (S i Place the article with the highest ) at the top, and the recommendation score (S i A list is generated such that the article with the lowest value is positioned at the bottom, and the dispatch recommendation unit collects the rating, distance, and response time of each article in real time and inputs them as input values to the article recommendation list generation algorithm. The dispatch recommendation unit utilizes the article recommendation list generation algorithm to perform normalization of the rating using a logarithmic function, transformation of the distance using a trigonometric function cosine, and normalization of the response time for all articles. The dispatch recommendation unit normalizes the variables of each element through the article recommendation list generation algorithm, then applies and sums the importance weight (w1) of the rating, the importance weight (w2) of the distance, and the importance weight (w3) of the response time. The dispatch recommendation unit calculates the recommendation score (S) for each article derived from the article recommendation list generation algorithm. i A list of recommended articles is generated by sorting in descending order of ) and the dispatch recommendation unit generates a list of recommended articles whenever a dispatch request is input from the client terminal, and the cumulative rating (R) of the corresponding article i ), distance between current departure point information and the corresponding article (D i ), average response time of the article (T i ) is collected and derived, and the dispatch recommendation unit is a driver-specific recommendation score (S) that is updated whenever a dispatch request is input from the client terminal. i A dispatch management system that generates a list of recommended articles based on ), and wherein the dispatch recommendation unit decreases the importance weight (w1) of the rating and increases the importance weight (w3) of the response time when there is a surge in delivery request inputs from the client terminal. Claim 2 delete Claim 3 delete Claim 4 delete Claim 5 delete Claim 6 delete
Citation Information
Patent Citations
Method, device and system for allocating delivery driver based on artificial intelligence
KR102499234B1
Apparatus and method for providing delivery services of agency and consignment based on artificial intelligence chatbot
KR102769475B1
Delivery time-based rate setting and calculation system
KR1020180101912A
Server, method and program for automatically adjusting fares based on artificial intelligence
KR102518979B1
Delivery logistics processing method, device and system based on analysis results of delivery driver's work status and logistics movement line
KR102749707B1
