Online car-hailing transport capacity prediction method and device, computer equipment and storage medium

Abstract

The invention relates to the technical field of online car-hailing, and discloses an online car-hailing transport capacity prediction method and device, computer equipment and a storage medium, and the method comprises the steps: dividing a target geographic region in an electronic map into a plurality of hexagonal grid units; acquiring real-time state data of each vehicle in each hexagonal grid unit, and determining real-time transport capacity data of each hexagonal grid unit according to the real-time state data; and predicting the transport capacity demand change condition of each hexagonal grid unit in a future time period based on the real-time transport capacity data. According to the method, the problems of inaccurate scheduling and response lag caused by insufficient grid division precision and lack of prediction capability in the existing online car-hailing transport capacity management are solved.

Description

Technical Field

[0001] This invention relates to the field of ride-hailing technology, specifically to a method, apparatus, computer equipment, and storage medium for predicting ride-hailing capacity. Background Technology

[0002] In the ride-hailing service sector, real-time monitoring of the distribution and status of transportation capacity resources across different areas of a city is crucial for efficient dispatching and optimized resource allocation. Currently, ride-hailing systems typically rely on GPS positioning technology to obtain driver location information and monitor and dispatch transportation capacity accordingly. However, due to the dynamic and uneven distribution of urban traffic flow, traditional methods often struggle to achieve precise characterization and dynamic response to transportation capacity status. This leads to issues such as idle drivers or insufficient capacity in some areas, impacting overall service efficiency and user experience.

[0003] In existing technologies, there are solutions for capacity management using geographic grids, but they still generally have two shortcomings: First, the grid division is mostly done in a rectangular manner, which is difficult to adapt to the spatial distribution characteristics of complex traffic flows, resulting in low accuracy of capacity statistics; second, the system focuses more on real-time status monitoring and lacks the ability to predict future capacity demand, making it difficult to respond to supply and demand fluctuations in a timely manner and to achieve forward-looking capacity scheduling. Summary of the Invention

[0004] In view of this, embodiments of the present invention provide a method, apparatus, computer equipment and storage medium for predicting ride-hailing capacity, in order to solve the problems of inaccurate scheduling and delayed response caused by insufficient grid division accuracy and lack of predictive ability in existing ride-hailing capacity management.

[0005] In a first aspect, embodiments of the present invention provide a method for predicting ride-hailing capacity, the method comprising: The target geographic area in the electronic map is divided into multiple hexagonal grid units; The real-time status data of each vehicle in each hexagonal grid cell is obtained, and the real-time capacity data of each hexagonal grid cell is determined based on the real-time status data. Based on the real-time capacity data, the changes in capacity demand for each of the hexagonal grid cells are predicted over a future time period.

[0006] Furthermore, determining the real-time transport capacity data of each hexagonal grid cell based on the real-time status data includes: Analyze the operational status of each vehicle in the real-time status data; The number of vehicles in an idle state within the hexagonal grid cell is calculated based on the operational status. The real-time transport capacity data for each hexagonal grid cell is determined based on the number of vehicles.

[0007] Furthermore, the step of predicting the change in capacity demand for each of the hexagonal grid cells within a future time period based on the real-time capacity data includes: Acquire external environmental data of the target geographic area, wherein the external environmental data includes at least weather information, traffic event information, and time period information; Generate corresponding real-time capacity features based on the external environment data and the real-time capacity data; Based on the real-time capacity characteristics, predict the changes in capacity demand for each of the hexagonal grid cells within the future time period.

[0008] Furthermore, the prediction of the capacity demand changes of each hexagonal grid cell within the future time period based on the real-time capacity characteristics includes: Obtain the historical transport capacity characteristics of the hexagonal grid cells within a historical time period; A time-series analysis is performed on the historical capacity characteristics and the real-time capacity characteristics to obtain the analysis results; Based on the analysis results, the changes in transport capacity demand for each hexagonal grid cell within the future time period are calculated.

[0009] Furthermore, after predicting the changes in capacity demand for each of the hexagonal grid cells over a future time period based on the real-time capacity data, the method further includes: The display parameters for each hexagonal grid cell are determined based on the real-time capacity data and the changes in capacity demand. Generate a visual identifier for each hexagonal grid cell according to the display parameters, and display the visual identifier on the electronic map of the user interface.

[0010] Furthermore, the method also includes: Responding to a user's triggering action on a target visualization marker on the electronic map; Based on the triggering operation, identify the target hexagonal grid cell corresponding to the target visualization identifier, and obtain the target capacity demand change corresponding to the target hexagonal grid cell; Generate a capacity demand change trend chart based on the changes in the target capacity demand, and display the capacity demand change trend chart on the user interface.

[0011] Furthermore, the method also includes: Based on the changes in transport capacity demand, generate transport capacity scheduling suggestions; The capacity scheduling suggestion is sent to the scheduling terminal, and the scheduling adjustment instruction is received from the scheduling terminal; The capacity allocation strategy for the corresponding hexagonal grid cell is updated according to the scheduling adjustment instruction.

[0012] Secondly, embodiments of the present invention provide a ride-hailing capacity prediction device, the device comprising: The partitioning module is used to divide the target geographic area in the electronic map into multiple hexagonal grid cells; The acquisition module is used to acquire the real-time status data of each vehicle in each hexagonal grid cell, and determine the real-time transport capacity data of each hexagonal grid cell based on the real-time status data. The prediction module is used to predict the changes in capacity demand for each of the hexagonal grid cells in the future time period based on the real-time capacity data.

[0013] Thirdly, embodiments of the present invention provide a computer device, including: a memory and a processor, the memory and the processor being communicatively connected to each other, the memory storing computer instructions, and the processor executing the computer instructions to perform the method described in the first aspect or any corresponding embodiment thereof.

[0014] Fourthly, embodiments of the present invention provide a computer-readable storage medium storing computer instructions that cause a computer to perform the method described in the first aspect or any of its corresponding embodiments.

[0015] The method provided in this application has the following beneficial effects: The method provided in this application divides the target geographical area in the electronic map into multiple hexagonal grid cells, making the area division more consistent with the spatial distribution of the actual road network and traffic flow, improving the accuracy and rationality of geospatial analysis, and providing a more accurate unit basis for capacity calculation. By acquiring the real-time status data of each vehicle in each hexagonal grid cell and determining the real-time capacity data accordingly, it achieves refined and dynamic monitoring of capacity distribution, enhancing the real-time performance and accuracy of capacity status judgment. Based on the real-time capacity data, it predicts the future changes in capacity demand of each grid cell, enabling the system to proactively identify supply and demand trends, providing decision support for capacity scheduling, thereby improving overall scheduling efficiency and service response capabilities. Attached Figure Description

[0016] To more clearly illustrate the specific embodiments of the present invention or the technical solutions in the prior art, the drawings used in the description of the specific embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are some embodiments of the present invention. For those skilled in the art, other drawings can be obtained from these drawings without creative effort.

[0017] Figure 1 This is a flowchart illustrating the ride-hailing capacity prediction method according to an embodiment of the present invention; Figure 2 This is a schematic diagram of hexagonal grid cell division in an electronic map according to an embodiment of the present invention; Figure 3 This is a flowchart illustrating another method for predicting ride-hailing capacity according to an embodiment of the present invention; Figure 4 This is a schematic diagram of the visual identifier of a grid cell on an electronic map according to an embodiment of the present invention; Figure 5 This is a flowchart illustrating another method for predicting ride-hailing capacity according to an embodiment of the present invention; Figure 6 This is a schematic diagram of the details card of an individual transportation capacity in an electronic map according to an embodiment of the present invention; Figure 7 This is a flowchart illustrating another method for predicting ride-hailing capacity according to an embodiment of the present invention. Figure 8 This is a schematic diagram of the workflow of the ride-hailing capacity prediction system according to an embodiment of the present invention; Figure 9 This is a structural block diagram of a ride-hailing capacity prediction device according to an embodiment of the present invention; Figure 10 This is a schematic diagram of the hardware structure of a computer device according to an embodiment of the present invention. Detailed Implementation

[0018] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0019] According to embodiments of the present invention, a method, apparatus, computer device, and storage medium for predicting ride-hailing capacity are provided. It should be noted that the steps shown in the flowcharts in the accompanying drawings can be executed in a computer system such as a set of computer-executable instructions. Furthermore, although a logical order is shown in the flowcharts, in some cases, the steps shown or described may be executed in a different order than that shown here.

[0020] This embodiment provides a method for predicting ride-hailing capacity. Figure 1 This is a flowchart of a ride-hailing capacity prediction method according to an embodiment of the present invention, such as... Figure 1 As shown, the process includes the following steps: Step S101: Divide the target geographic area in the electronic map into multiple hexagonal grid cells.

[0021] In this embodiment, the electronic map refers to geographic information data stored and displayed in digital form, which includes spatial information such as road networks, points of interest, and administrative divisions. The target geographic area refers to a specific geographic range that needs to be monitored and predicted for transportation capacity, such as the entire jurisdiction of a city, a business district, or the area surrounding a transportation hub. The implementation method is as follows: based on the H3 hexagonal spatial indexing system, the target geographic area is automatically divided into seamlessly joined hexagonal grid cells of equal area within the coordinate system of the electronic map. Each hexagonal grid cell, as an independent honeycomb spatial unit, has geometric characteristics that, compared to traditional rectangular grids, better conform to the natural extension of the urban road network and the actual distribution of traffic flow, thus providing a geographic reference basis for subsequent transportation capacity calculations at the spatial division level. In specific implementation, different resolution levels of the H3 system can be selected according to the management accuracy requirements to generate grid cells of appropriate size. For example, a higher resolution (smaller hexagons) can be used in the city center to achieve refined analysis, while a lower resolution (larger hexagons) can be used in the suburbs to reduce computational complexity. This partitioning process can be static preprocessing or dynamically executed based on regional boundaries, ensuring that the entire target geographic area is completely and non-overlappingly covered, laying a unified spatial framework for subsequent capacity data collection, aggregation, and visualization based on grid units.

[0022] As an example, such as Figure 2 As shown, the above partitioning process is implemented using the H3 grid generation algorithm, which transforms the city's geographic coordinate data into a structured set of hexagonal grids. Due to its geometric characteristics, the hexagonal grid is better suited to the spatial distribution of complex urban traffic flows compared to traditional rectangular grids, reducing geographic errors from the data source and laying the foundation for establishing a high-precision geographic index of transportation capacity.

[0023] Step S102: Obtain the real-time status data of each vehicle in each hexagonal grid cell, and determine the real-time transport capacity data of each hexagonal grid cell based on the real-time status data.

[0024] In this embodiment, by integrating a GPS positioning module, vehicle sensors, and mobile application interfaces, multiple data sources are continuously collected for each vehicle, including latitude and longitude coordinates, instantaneous speed, order acceptance status, and vehicle health indicators, forming a high-frequency real-time status data stream. This data is transmitted to the cloud via 5G or vehicle-to-everything (V2X) communication and then processed in real-time using a stream processing framework (such as Apache RocketMQ) for deduplication and spatiotemporal alignment. Based on the analysis results, the operational status of each vehicle (e.g., idle, in service, suspended service) is identified. Then, using hexagonal grid cells as spatial aggregation units, the number of idle vehicles within each cell is accumulated, and this number is directly used as the core indicator of real-time capacity data. Capacity density is calculated by combining the grid cell area, or weighting factors such as vehicle passenger capacity and driver service ratings are introduced to generate a weighted capacity value. Real-time status data can be acquired by integrating vehicle trajectory interpolation technology to smooth out GPS positioning jitter, or by accessing real-time road traffic speed data to help determine vehicle availability; real-time capacity data can also be determined by online statistical algorithms, such as the average number of idle vehicles within a sliding time window, or by fusing and inferring multi-source status features based on a lightweight machine learning model to output a more robust capacity estimate.

[0025] As an example, for a hexagonal grid cell identified as "H3-8a2b3c", 15 vehicle status data packets were received in the most recent collection cycle (e.g., within 30 seconds). After parsing, it was determined that 9 vehicles were idle, 4 were in service, and 2 were offline. The number of idle vehicles in this cell was counted as 9, and this was recorded as the current real-time capacity data. If the area of ​​this cell at H3 resolution is 0.5 square kilometers, the capacity density is calculated to be 18 vehicles / square kilometer, which is then used by the subsequent scheduling and prediction modules.

[0026] Step S103: Based on real-time capacity data, predict the changes in capacity demand for each hexagonal grid cell in the future time period.

[0027] In this embodiment, the latest real-time capacity data (such as the current number of idle vehicles) and its short-term historical sequence of each grid cell are used as input features. A pre-trained time-series prediction model (such as a Long Short-Term Memory network, an autoregressive ensemble moving average model, or a Transformer-based sequence model) is invoked. This model learns the inherent time dependence and trend patterns of the data to directly deduce the predicted value sequence of capacity demand for the grid cell at equal future time points, or to calculate the rate of change of demand relative to the current level. In addition, the prediction process can also be designed as an online streaming prediction, that is, rolling prediction is triggered whenever new real-time capacity data is received to ensure the timeliness of the prediction results; a multi-scale prediction framework can also be introduced to simultaneously generate short-term (such as the next 15 minutes) and medium-term (such as the next 1 hour) demand change curves to support scheduling decisions for different response periods.

[0028] As an example, for a grid cell identified as "H3-8a2b3c", its current real-time capacity data is obtained as 12 idle vehicles, and its data sequence of the past 4 periods (one period every 5 minutes) [10, 11, 10, 12] vehicles is loaded as supplementary features; input into the deployed gradient boosting tree regression model, the model outputs the capacity demand prediction sequence for the next 3 periods (i.e. the next 15 minutes) as [13, 15, 14] vehicles; this sequence is recorded as the capacity demand change of this cell in the future time period, indicating that the demand is expected to rise first and then fall slightly, providing a quantitative basis for pre-scheduling.

[0029] In this embodiment of the application, determining the real-time capacity data of each hexagonal grid cell based on real-time status data includes: Step A1: Analyze the operational status of each vehicle in the real-time status data.

[0030] Specifically, real-time status data refers to the dynamic vehicle information data stream collected and uploaded in real time through vehicle-to-everything (V2X) terminals, GPS devices, and mobile applications. It is typically transmitted in the form of structured data packets, and its data fields include at least vehicle identifier, location coordinates, instantaneous speed, order acceptance status identifier, and vehicle health status code. Operational status is a key attribute characterizing whether a vehicle is currently available for dispatching and order acceptance, mainly including discrete status categories such as idle, in service, and suspended service. The implementation method is as follows: Real-time status data streams from various vehicles are received; first, they are deserialized and parsed according to a preset data protocol to extract key field values; then, logical judgments are made on these fields according to predefined status determination rules to map the specific operational status.

[0031] For example, if the order status identifier field value is "idle" and the instantaneous speed is below a threshold (e.g., 5 km / h), while the vehicle health status code indicates normal operation, then the vehicle's operational status is determined to be "idle." If the order status identifier is "serviced," then regardless of the speed, it is determined to be "serviced." If the vehicle health status code indicates abnormality or the order status identifier is "offline," then it is determined to be "service suspended." This parsing process can be implemented using a rule engine or state machine to ensure efficient and accurate parsing of massive amounts of real-time data, providing a foundation for subsequent capacity aggregation based on operational status.

[0032] Step A2: Calculate the number of vehicles in an idle state within the hexagonal grid cell based on the operational status.

[0033] Specifically, an in-memory data structure (such as a hash map) is maintained, using unique hexagonal grid cell identifiers as keys and vehicle status lists as values. After parsing a batch of real-time vehicle status data, the hexagonal grid cell identifier to which each vehicle belongs is quickly located using the H3 spatial indexing function based on the real-time location coordinates reported by each vehicle. The vehicle identifier is added to the corresponding grid cell's count set only when the vehicle's operating status is determined to be idle. Simultaneously, the contents of the sets within each grid cell are dynamically updated by monitoring vehicle status change events in real time (such as transitioning from idle to active), ensuring the real-time nature of the statistics. In each statistical cycle (such as every 30 seconds), all grid cells are traversed, and the number of unique vehicle identifiers within each cell set is directly calculated, thus obtaining the accurate number of vehicles in the idle state within that cell. This process can be implemented using a distributed counting service, supporting high-concurrency updates and queries. Furthermore, time window filtering can be set to avoid count fluctuations caused by vehicles briefly entering and leaving the grid, thereby providing a stable and reliable input for subsequent capacity data generation.

[0034] Step A3: Determine the real-time transport capacity data for each hexagonal grid cell based on the number of vehicles.

[0035] Specifically, real-time capacity data is a core indicator used to quantify the availability of capacity in a grid cell at the current moment. This is implemented by directly using the number of idle vehicles in each grid cell as the basic capacity value. Simultaneously, to enhance data comparability and interpretability, the capacity density per unit area (e.g., number of idle vehicles per square kilometer) can be calculated by combining the area of ​​the hexagonal grid cell (pre-calculated by the H3 indexing system), and this density value is used as more refined real-time capacity data. Furthermore, to achieve stable data output and rapid querying, the capacity data (basic quantity or density value), along with the timestamp and grid cell identifier, can be encapsulated into a data structure and written in real-time to a high-performance in-memory database (such as Redis) or a time-series database for subsequent prediction and visualization modules to subscribe to and access. This step ensures the transformation from raw counts to standardized capacity indicators, providing the system with a spatiotemporal data foundation that can be directly used for analysis and decision-making.

[0036] In this embodiment of the application, predicting the change in capacity demand for each hexagonal grid cell within a future time period based on real-time capacity data includes: Step B1: Obtain external environmental data for the target geographic area, including at least weather information, traffic event information, and time period information.

[0037] Specifically, by calling third-party data service APIs or accessing relevant data streams, three types of key information associated with the target geographic area (i.e., the overall area covered by the pre-defined hexagonal grid cells) are collected in real-time or near real-time. Weather information is obtained through accessing the forecast data interface of meteorological departments, covering indicators such as precipitation, temperature, and visibility in the area where each grid cell is located; traffic event information is obtained by integrating event release platforms of traffic management departments or user-reported data, acquiring real-time traffic status such as accidents, traffic control, and congestion; time period information is automatically generated by the local clock, including date type (weekday / holiday) and specific time periods of the day (such as morning rush hour, evening rush hour). This data is parsed and spatially aligned, associated with the corresponding hexagonal grid cells, and attached with a valid timestamp, forming a structured external environment dataset, providing multi-dimensional input for subsequent feature generation.

[0038] Step B2: Generate corresponding real-time capacity characteristics based on external environment data and real-time capacity data.

[0039] Specifically, for each grid cell, its corresponding external environmental data is first standardized and encoded. For example, weather information (such as "light rain") is mapped to a predefined category code, traffic event information (such as "accident occurred") is converted into Boolean symbols, and time period information (such as "18:00-19:00") is converted into a periodic time point within a day. Simultaneously, real-time capacity data (such as the number of idle vehicles or capacity density) is normalized to eliminate the influence of dimensions. These encoded and normalized multi-dimensional data fields are then concatenated into a fixed-length numerical vector according to a predetermined feature order. This vector constitutes the real-time capacity feature of the grid cell at the current moment. This feature generation process can be implemented through streaming computing tasks, ensuring that the latest feature vector is output for each grid cell in each update cycle, providing a directly input data foundation for subsequent time series analysis and prediction.

[0040] Step B3: Based on real-time capacity characteristics, predict the changes in capacity demand for each hexagonal grid cell over a future time period.

[0041] Specifically, based on real-time capacity characteristics, the forecast of capacity demand changes for each hexagonal grid cell over a future time period includes: Step B301: Obtain the historical transport capacity characteristics of the hexagonal grid cells within the historical time period.

[0042] Specifically, historical capacity features refer to the pre-generated and stored time-series feature data set corresponding to each hexagonal grid cell within a historical time period. Its feature structure is consistent with real-time capacity features, including historical capacity values ​​and corresponding historical external environment data encoding. Accessing the time-series database or distributed file system (such as ElasticSearch) used to store historical feature data, based on the historical time period defined by the current forecast task (e.g., data from the same grid cell within the same time period over the past 30 days), and using the grid cell identifier and timestamp as the joint query key, all feature vector sequences of that cell within the same historical time period can be retrieved in batches. These historical features have undergone the same standardization and encoding process during storage, ensuring spatial and semantic alignment with real-time features. In practical implementation, the length and granularity of the historical time period can be dynamically adjusted according to forecasting needs. For example, to forecast capacity demand during weekday morning rush hours, data from the same morning rush hour over the past four weekdays can be extracted as a historical feature set, thus providing a timely and periodic comparative benchmark for subsequent time-series analysis.

[0043] Step B302 involves performing a time-series analysis on historical and real-time capacity characteristics to obtain the analysis results.

[0044] Specifically, time series analysis refers to the mathematical and statistical algorithmic process aimed at extracting patterns, trends, and periodicities from time-series characteristic data. The real-time capacity characteristics of the current moment are treated as a point-in-time data point and concatenated with historical capacity characteristic sequences sorted by timestamps to form a complete time series. Subsequently, time series forecasting algorithms (such as Autoregressive Integral Moving Average (ARIMA), Long Short-Term Memory (LSTM), or Prophet models) are used to model and analyze this series. This analysis process not only considers the numerical values ​​of the series itself but also automatically identifies and quantifies trend components (such as long-term growth or decline in capacity demand), seasonal or periodic components (such as daily peaks, weekly patterns), and possible sudden fluctuations. The model output is not a direct predicted value but rather an intermediate state vector or model parameters. These intermediate products are collectively referred to as the analysis results of this step, which carry the time series evolution patterns mined from historical to current data, providing direct input for the next step of specific prediction calculations. Furthermore, multiple time series models based on different principles can be run in parallel, and their analysis results can be fused through ensemble learning to improve the robustness and accuracy of the analysis.

[0045] Step B303: Calculate the changes in transport capacity demand for each hexagonal grid cell over the future time period based on the analysis results.

[0046] Specifically, based on the analysis results (i.e., intermediate outputs of the model such as time-series patterns, trends, and periodic characteristics extracted from historical and real-time data), the capacity demand changes for each hexagonal grid cell within the future time period are calculated. The capacity demand changes are a quantitative predictive output, typically represented as a sequence of predicted values ​​or a rate-of-change curve of the required capacity (e.g., the number of available vehicles) for that grid cell at consecutive future time points (e.g., every 5-minute interval within the next 30 minutes). The implementation involves using the analysis results as input and loading a pre-trained predictive model (e.g., a time series model, regression model, or neural network). This model directly calculates the capacity demand value for each future time point based on the time-series patterns implied in the analysis results. For example, if the analysis results contain parameters of an ARIMA model or hidden states of an LSTM network, the model can perform forward computation based on these parameters or states to generate a predicted sequence for the future time period. In addition, the calculation process can also incorporate multi-model ensemble, that is, combine the prediction results of multiple independent models and obtain the final prediction through weighted averaging or stacking methods to improve accuracy; at the same time, the prediction range with confidence interval can be output for each grid cell, thereby quantifying the uncertainty of the prediction and providing a more comprehensive reference for scheduling decisions.

[0047] In this embodiment of the application, after predicting the changes in capacity demand for each hexagonal grid cell over a future time period based on real-time capacity data, as follows: Figure 3 As shown, the method also includes: Step S201: Determine the display parameters for each hexagonal grid cell based on real-time capacity data and changes in capacity demand.

[0048] In this embodiment, real-time capacity data (such as the current number of available vehicles) is first normalized and converted into basic visual variable values ​​based on preset mapping rules (e.g., mapping capacity values ​​to saturation values ​​between 0 and 1 or hue values ​​from light red to dark red). Simultaneously, this basic visual variable is dynamically adjusted based on predicted changes in capacity demand for the unit over future time periods (e.g., the rate of change of the predicted demand value for the next 15 minutes compared to the current value). For example, if a significant increase in demand is predicted (supply shortage trend), a flashing effect or a highlighted outer ring can be overlaid on the basic color as an enhanced display parameter to alert operators; if supply and demand are predicted to be balanced, the basic color is maintained. Finally, a structured display parameter object is generated for each grid cell, containing at least key-value pairs such as color code, transparency, whether dynamic effects are enabled, and their frequency. This process can be implemented through an independent visualization configuration service, supporting flexible strategy configuration, thereby transforming abstract spatiotemporal data into an intuitive visual language that guides decision-making.

[0049] Step S202: Generate a visual identifier corresponding to each hexagonal grid cell according to the display parameters, and display the visual identifier on the electronic map of the user interface.

[0050] In this embodiment, based on the display parameters generated for each hexagonal grid cell, a corresponding visual identifier is dynamically created and rendered, and then overlaid on the electronic map of the user interface. The visual identifier refers to a graphic overlay (e.g., a hexagonal face filled with a specific color) generated according to the visual attributes (such as color, transparency, and dynamic effects) defined by the display parameters. The front-end visualization engine (such as a WebGL-based map rendering framework) receives a display parameter data stream from a back-end service. This data stream is indexed by the grid cell identifier and contains the corresponding visual attributes. The engine then uses these parameters to draw a polygonal layer at the corresponding geographic coordinates on the electronic map, precisely matching the grid cell boundaries and possessing a specified fill color (such as a gradient from light red to dark red) and transparency. If dynamic effects (such as blinking or highlighting) are enabled in the display parameters, the engine synchronously attaches corresponding animation instructions. In addition, layered rendering and view clipping techniques can be used to render only the grid cells within the current map's visible range, and spatial indexes can be used to manage massive amounts of labels, ensuring a smooth user experience when interacting with the map (such as panning and zooming), thereby transforming abstract capacity data into visual language on the map in real time and intuitively.

[0051] As an example, such as Figure 4As shown, the generated visual markers are presented as a series of colored hexagons on the electronic map of the user interface. A color gradient from light red to dark red is used to map real-time capacity data for different grid cells: lighter colors indicate tighter available capacity within the cell, while darker colors indicate more abundant capacity. This design allows operators to instantly perceive the distribution of available capacity across the entire target geographical area through a visual interface, achieving macro-level, real-time monitoring of capacity status and greatly improving situational awareness and decision-making efficiency.

[0052] In the embodiments of this application, such as Figure 5 As shown, the method also includes: Step S301, in response to the user's triggering operation on the target visualization mark on the electronic map.

[0053] In this embodiment, a target visual identifier specifically refers to a graphic element (i.e., a rendered specific hexagonal graphic) that a user sees and intends to interact with on an electronic map, representing a specific hexagonal grid cell. A trigger operation refers to a clear interactive command issued by the user to the target visual identifier via an input device (such as a mouse or touchscreen), such as a mouse click, finger tap, or touchscreen long press. When rendering each visual identifier (hexagonal graphic), the front-end visualization engine binds a specific event listener (such as a click or tap event) to it. When a user performs a trigger operation on an identifier on the electronic map, the browser or mobile framework generates an interaction event object, which contains metadata such as event type and screen coordinates. After capturing the event, the event listener determines the target visual identifier through coordinate transformation and graphic picking algorithms. Based on the unique grid cell identifier embedded or associated with the identifier during rendering, the target hexagonal grid cell of interest to the user can be immediately identified, thus providing clear input for subsequent steps (obtaining prediction data for the cell and generating a trend chart).

[0054] Step S302: Based on the trigger operation, identify the target hexagonal grid cell corresponding to the target visualization identifier, and obtain the target capacity demand change corresponding to the target hexagonal grid cell.

[0055] In this embodiment, based on the identified target hexagonal grid cell (i.e., the specific grid cell specified by the user through a trigger operation), the target capacity demand change information corresponding to that cell is retrieved from the background data service. The target capacity demand change information refers to the calculated and stored future time-period capacity demand forecast results (such as a sequence of predicted values) specific to that target grid cell. After determining the unique identifier of the target grid cell, the front-end application immediately initiates an asynchronous request (such as an AJAX or gRPC call) to the background forecast data service, carrying the identifier and the required forecast time range in the request parameters. Upon receiving the request, the background service uses the identifier as the key to query the pre-stored forecast result data from a cache (such as Redis) or time-series database, and encapsulates it into a structured response (such as JSON format, containing forecast values ​​and timestamps for multiple future time points) and returns it to the front-end. If the forecast data is not ready due to calculation delays, it can return to the calculation status and start polling, pushing the data to the front-end when it becomes available; or, based on the interaction context of the user interface, it can synchronously return richer related data for that cell, such as real-time capacity values ​​and historical comparisons, providing a more comprehensive data foundation for generating trend charts.

[0056] Step S303: Generate a capacity demand change trend chart based on the changes in target capacity demand, and display the capacity demand change trend chart on the user interface.

[0057] In this embodiment, a corresponding capacity demand trend chart is generated based on the changes in target capacity demand (i.e., the predicted sequence of future capacity demand for a specific target grid cell), and rendered and displayed on the user interface. The capacity demand trend chart is a statistical chart (usually in the form of a line chart) with time as the horizontal axis and the predicted capacity demand value (such as the required number of available vehicles) as the vertical axis, used to visually display the demand trend of the region over a future period. After receiving the structured prediction data (containing multiple future time points and their corresponding predicted values), the front-end application calls an integrated chart rendering library (such as ECharts or D3.js) to map the data into the required coordinate point sequence for the chart; a floating layer or side panel is created in the user interface, and the completed trend chart is embedded within it. This chart typically includes elements such as coordinate axes, data lines, and key point markers, and can be supplemented with auxiliary information such as prediction confidence interval shading. In addition, the charts support interactive operations, such as hovering the mouse to display specific values, zooming the time axis to view more detailed or longer-term trends, or providing comparison functions to overlay historical data from the same period, thereby providing users with in-depth analysis capabilities and helping them understand the dynamic patterns of capacity changes.

[0058] Furthermore, visualization and interactivity are not limited to regional-level trend analysis. For example... Figure 6As shown, during the visualized monitoring of transport capacity status, users can also click on the coordinates of specific drivers or vehicles on the electronic map to trigger the display of a detailed card for that individual transport capacity. This card intuitively displays micro-level information such as the driver's name, dispatch status, service status, and order preferences, and provides access to deeper operations such as viewing routes and details. This achieves penetrating query and control from the macro-grid regional situation to the micro-individual status, enabling operators to grasp the overall distribution while accurately confirming the status and intervening in the scheduling of specific transport capacity resources, thus improving the precision of scheduling decisions and the operational flexibility of the system.

[0059] In the embodiments of this application, such as Figure 7 As shown, the method also includes: Step S401: Generate capacity scheduling suggestions based on changes in capacity demand.

[0060] In the embodiments of this application, the capacity scheduling suggestion refers to a set of operable scheduling instructions or strategy prompts that are automatically generated based on the predicted changes in capacity demand of each hexagonal grid cell (i.e., the predicted value or trend of capacity demand in a specific future time period) and through preset scheduling rules and optimization algorithms.

[0061] First, the predicted results are compared with real-time capacity data to identify grid units where supply and demand imbalances may occur in the future (e.g., hotspot areas where predicted demand is significantly higher than current available capacity, or coldspot areas with excess capacity). Then, based on a pre-set scheduling rule base (e.g., shortest distance, highest balance coefficient) and optimization objectives (e.g., minimizing overall empty driving distance, maximizing driver order-accepting efficiency), specific suggested measures are calculated and generated for each target unit. These measures may include: dispatching X idle vehicles from grid unit A to grid unit B, suggesting recruitment of surrounding capacity in grid unit C, or sending premium incentive prompts to drivers in grid unit D. The generation process can employ operations research models (e.g., linear programming or network flow models) for global optimization calculations, or introduce reinforcement learning models to dynamically adjust rule weights, ensuring that scheduling suggestions not only respond to immediate gaps but also balance long-term operational efficiency and driver income.

[0062] Step S402: Send the capacity scheduling suggestion to the scheduling terminal and receive the scheduling adjustment instruction from the scheduling terminal.

[0063] In this embodiment, the dispatch terminal refers to dedicated software or a web interface used by operations dispatchers to receive system suggestions and engage in decision-making interactions. Structured capacity dispatch suggestions (including fields such as suggestion type, source / target grid unit, number of vehicles, and suggestion execution time window) are sent to logged-in dispatch terminals that have subscribed to the relevant areas via push notifications or through internal message middleware (such as Apache RocketMQ) or real-time communication protocols (such as WebSocket). The terminal interface refreshes in real time, clearly displaying the suggestion content in the form of lists, pop-ups, or map annotations. Dispatchers can review, modify (e.g., adjust the number of dispatched vehicles), or reject suggestions on the terminal based on their professional judgment and submit confirmation. The terminal then encapsulates the human decision-making result into a structured dispatch adjustment instruction (including the original suggestion ID, adjusted parameters, operator identifier, timestamp, etc.) and sends it back to the system backend via API. Furthermore, two-way authentication can be used to ensure the security of the instruction source, and the format and business logic of the returned instructions can be verified (e.g., verifying whether the number of dispatched vehicles exceeds the actual available number) to ensure that the received instructions are legal and valid, laying a reliable input foundation for subsequent strategy updates.

[0064] Step S403: Update the capacity allocation strategy of the corresponding hexagonal grid cell according to the scheduling adjustment instruction.

[0065] In this embodiment, the capacity allocation strategy refers to a set of dynamic rules or parameters (e.g., the weight of idle vehicles to be dispatched, dispatch priority, recommended dispatch radius, etc.) used within the system to guide the automatic dispatch algorithm. The system parses the instruction type (e.g., modifying the dispatch quantity or rejecting a dispatch suggestion), the target grid cell identifier, and the adjusted parameters contained in the dispatch adjustment instruction. Then, it accesses the database or configuration center storing the strategy configuration, locates the current strategy record using the target grid cell as an index, and atomically updates the adjusted content in the instruction (e.g., updating the number of automatically dispatched suggested vehicles from 5 to 3) to that record, thus completing the strategy revision. The update operation can be completed within a database transaction, generating a strategy change log. After a successful update, a strategy refresh notification is immediately sent to the relevant real-time dispatch calculation module, enabling subsequent automatic dispatch decisions to be executed immediately based on the new strategy, thereby achieving a closed-loop fusion of human experience and system intelligence.

[0066] This embodiment provides a ride-hailing capacity prediction system, such as... Figure 8 As shown, the system includes: a real-time driver data acquisition module, a data processing and analysis module, a capacity status prediction module, and a visualization module; The real-time driver data acquisition module is used to acquire real-time status data of each vehicle in each hexagonal grid cell. Specifically, it includes real-time acquisition of driver location, order acceptance status, vehicle health and behavior data through vehicle GPS and mobile application interface, and reporting it to the system. The data processing and analysis module is used to divide the target geographical area in the electronic map into multiple hexagonal grid units and determine the real-time transportation capacity data of each hexagonal grid unit based on the real-time status data. Specifically, it includes: dividing the geographical grid based on the H3 algorithm, and performing streaming processing, parsing and aggregation on the reported real-time status data to calculate the real-time transportation capacity data of each hexagonal grid unit. The capacity status prediction module is used to predict the changes in capacity demand of each hexagonal grid cell in the future time period based on the real-time capacity data. Specifically, it includes: combining external environmental data and real-time capacity characteristics, and using time series analysis and machine learning models to predict the future changes in capacity demand of each grid cell. The visualization module is used to generate visual icons based on real-time capacity data and forecast results and render them on an electronic map. Users can view details, trend charts and capacity scheduling suggestions through interaction, enabling intuitive monitoring and scheduling intervention of capacity status.

[0067] This embodiment also provides a ride-hailing capacity prediction device, which is used to implement the above embodiments and preferred embodiments; details already described will not be repeated. As used below, the term "module" can be a combination of software and / or hardware that implements a predetermined function. Although the device described in the following embodiments is preferably implemented in software, hardware implementation, or a combination of software and hardware, is also possible and contemplated.

[0068] This embodiment provides a ride-hailing capacity prediction device, such as... Figure 9 As shown, it includes: The partitioning module 91 is used to divide the target geographic area in the electronic map into multiple hexagonal grid units; The acquisition module 92 is used to acquire the real-time status data of each vehicle in each hexagonal grid cell, and determine the real-time capacity data of each hexagonal grid cell based on the real-time status data. The prediction module 93 is used to predict the changes in capacity demand for each hexagonal grid cell in the future time period based on real-time capacity data.

[0069] In this embodiment of the application, the acquisition module 92 is specifically used to parse the operating status of each vehicle in the real-time status data; calculate the number of vehicles in the idle state within the hexagonal grid cell based on the operating status; and determine the real-time transport capacity data of each hexagonal grid cell based on the number of vehicles.

[0070] In this embodiment of the application, the prediction module 93 is specifically used to acquire external environmental data of the target geographical area, wherein the external environmental data includes at least weather information, traffic event information and time period information; generate corresponding real-time transportation capacity features based on the external environmental data and real-time transportation capacity data; and predict the change in transportation capacity demand of each hexagonal grid cell in the future time period based on the real-time transportation capacity features.

[0071] In this embodiment of the application, the prediction module 93 is specifically used to obtain the historical capacity characteristics of the hexagonal grid cells within a historical time period; perform time series analysis on the historical capacity characteristics and real-time capacity characteristics to obtain the analysis results; and calculate the change in capacity demand of each hexagonal grid cell within a future time period based on the analysis results.

[0072] In this embodiment of the application, the device further includes: a first display module, used to determine the display parameters of each hexagonal grid cell based on real-time capacity data and changes in capacity demand; generate a visual identifier corresponding to each hexagonal grid cell according to the display parameters; and display the visual identifier on the electronic map of the user interface.

[0073] In this embodiment of the application, the device further includes: a second display module, configured to respond to a user's trigger operation on a target visual identifier on an electronic map; identify the target hexagonal grid cell corresponding to the target visual identifier based on the trigger operation, and obtain the target capacity demand change status corresponding to the target hexagonal grid cell; generate a capacity demand change trend map based on the target capacity demand change status, and display the capacity demand change trend map on the user interface.

[0074] In this embodiment of the application, the device further includes: an update module, configured to generate capacity scheduling suggestions based on changes in capacity demand; send the capacity scheduling suggestions to a scheduling terminal and receive scheduling adjustment instructions from the scheduling terminal; and update the capacity allocation strategy of the corresponding hexagonal grid cell according to the scheduling adjustment instructions.

[0075] Please see Figure 10 , Figure 10 This is a schematic diagram of the structure of a computer device provided in an optional embodiment of the present invention, such as... Figure 10As shown, the computer device includes one or more processors 10, memory 20, and interfaces for connecting the components, including high-speed interfaces and low-speed interfaces. The components communicate with each other via different buses and can be mounted on a common motherboard or otherwise installed as needed. The processors can process instructions executed within the computer device, including instructions stored in or on memory to display graphical information of a GUI on external input / output devices (such as display devices coupled to the interfaces). In some alternative implementations, multiple processors and / or multiple buses can be used with multiple memories and multiple memory modules, if desired. Similarly, multiple computer devices can be connected, each providing some of the necessary operations (e.g., as a server array, a group of blade servers, or a multiprocessor system).

[0076] Processor 10 may be a central processing unit, a network processor, or a combination thereof. Processor 10 may further include a hardware chip. The hardware chip may be an application-specific integrated circuit (ASIC), a programmable logic device (PLD), or a combination thereof. The programmable logic device may be a complex programmable logic device (CAMP), a field-programmable gate array (FPGA), a general-purpose array logic (GDA), or any combination thereof.

[0077] The memory 20 stores instructions executable by at least one processor 10 to cause at least one processor 10 to perform the method shown in the above embodiments.

[0078] The memory 20 may include a program storage area and a data storage area. The program storage area may store the operating system and applications required for at least one function; the data storage area may store data created based on the use of the computer device as shown by a landing page for an app. Furthermore, the memory 20 may include high-speed random access memory and may also include non-transitory memory, such as at least one disk storage device, flash memory device, or other non-transitory solid-state storage device. In some alternative embodiments, the memory 20 may optionally include memory remotely located relative to the processor 10, which can be connected to the computer device via a network. Examples of such networks include, but are not limited to, the Internet, intranets, local area networks, mobile communication networks, and combinations thereof.

[0079] The memory 20 may include volatile memory, such as random access memory; the memory may also include non-volatile memory, such as flash memory, hard disk or solid-state drive; the memory 20 may also include a combination of the above types of memory.

[0080] The computer device also includes a communication interface 30 for communicating with other devices or communication networks.

[0081] This invention also provides a computer-readable storage medium. The methods described above according to embodiments of the invention can be implemented in hardware or firmware, or implemented as computer code that can be recorded on a storage medium, or implemented as computer code downloaded via a network and originally stored on a remote storage medium or a non-transitory machine-readable storage medium and then stored on a local storage medium. Thus, the methods described herein can be processed by software stored on a storage medium using a general-purpose computer, a dedicated processor, or programmable or dedicated hardware. The storage medium can be a magnetic disk, optical disk, read-only memory, random access memory, flash memory, hard disk, or solid-state drive, etc.; further, the storage medium can also include combinations of the above types of memory. It is understood that computers, processors, microprocessor controllers, or programmable hardware include storage components capable of storing or receiving software or computer code, which, when accessed and executed by the computer, processor, or hardware, implements the methods shown in the above embodiments.

[0082] Although embodiments of the invention have been described in conjunction with the accompanying drawings, those skilled in the art can make various modifications and variations without departing from the spirit and scope of the invention, and such modifications and variations all fall within the scope defined by the appended claims.

Claims

1. A method for predicting ride-hailing capacity, characterized in that, The method includes: The target geographic area in the electronic map is divided into multiple hexagonal grid units; The real-time status data of each vehicle in each hexagonal grid cell is obtained, and the real-time capacity data of each hexagonal grid cell is determined based on the real-time status data. Based on the real-time capacity data, the changes in capacity demand for each of the hexagonal grid cells are predicted over a future time period.

2. The method according to claim 1, characterized in that, The step of determining the real-time transport capacity data of each hexagonal grid cell based on the real-time status data includes: Analyze the operational status of each vehicle in the real-time status data; The number of vehicles in an idle state within the hexagonal grid cell is calculated based on the operational status. The real-time transport capacity data for each hexagonal grid cell is determined based on the number of vehicles.

3. The method according to claim 1, characterized in that, The prediction of changes in capacity demand for each hexagonal grid cell within a future time period based on the real-time capacity data includes: Acquire external environmental data of the target geographic area, wherein the external environmental data includes at least weather information, traffic event information, and time period information; Generate corresponding real-time capacity features based on the external environment data and the real-time capacity data; Based on the real-time capacity characteristics, predict the changes in capacity demand for each of the hexagonal grid cells within the future time period.

4. The method according to claim 3, characterized in that, The prediction of the capacity demand changes of each hexagonal grid cell within the future time period based on the real-time capacity characteristics includes: Obtain the historical transport capacity characteristics of the hexagonal grid cells within a historical time period; A time-series analysis is performed on the historical capacity characteristics and the real-time capacity characteristics to obtain the analysis results; Based on the analysis results, the changes in transport capacity demand for each hexagonal grid cell within the future time period are calculated.

5. The method according to claim 1, characterized in that, After predicting the changes in capacity demand for each of the hexagonal grid cells over a future time period based on the real-time capacity data, the method further includes: The display parameters for each hexagonal grid cell are determined based on the real-time capacity data and the changes in capacity demand. Generate a visual identifier for each hexagonal grid cell according to the display parameters, and display the visual identifier on the electronic map of the user interface.

6. The method according to claim 5, characterized in that, The method further includes: Responding to a user's triggering action on a target visualization marker on the electronic map; Based on the triggering operation, identify the target hexagonal grid cell corresponding to the target visualization identifier, and obtain the target capacity demand change corresponding to the target hexagonal grid cell; Generate a capacity demand change trend chart based on the changes in the target capacity demand, and display the capacity demand change trend chart on the user interface.

7. The method according to claim 1, characterized in that, The method further includes: Based on the changes in transport capacity demand, generate transport capacity scheduling suggestions; The capacity scheduling suggestion is sent to the scheduling terminal, and the scheduling adjustment instruction is received from the scheduling terminal; The capacity allocation strategy for the corresponding hexagonal grid cell is updated according to the scheduling adjustment instruction.

8. A ride-hailing capacity prediction device, characterized in that, The device includes: The partitioning module is used to divide the target geographic area in the electronic map into multiple hexagonal grid cells; The acquisition module is used to acquire the real-time status data of each vehicle in each hexagonal grid cell, and determine the real-time transport capacity data of each hexagonal grid cell based on the real-time status data. The prediction module is used to predict the changes in capacity demand for each of the hexagonal grid cells in the future time period based on the real-time capacity data.

9. A computer device, characterized in that, include: A memory and a processor, the memory and the processor being communicatively connected to each other, the memory storing computer instructions, the processor executing the computer instructions to perform the method of any one of claims 1 to 7.

10. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores computer instructions for causing the computer to perform the method of any one of claims 1 to 7.

Images