Big data-based old-age care meal-assisting platform management method and system

Abstract

The invention relates to an old-age care meal-assisting platform management method and system based on big data, and the method comprises the following steps: collecting and storing multi-source data, the multi-source data comprising historical data and real-time data, the historical data comprising user data, order data and meal-assisting network data, and the real-time data comprising user data, order data and meal-assisting network data; the real-time data comprises external data, user behavior data and logistics data; based on the multi-source data, the big data is utilized to carry out meal assistance demand prediction, the meal assistance demand prediction comprises geographical distribution prediction, time distribution prediction, crowd characteristic distribution prediction and service type distribution prediction, and the prediction result is represented by the order quantity and the order type; according to the method, real-time order data is acquired, and under limited deliverymen and vehicle resources, a shortest path and an optimal task allocation scheme are designed by using a genetic algorithm to optimize delivery resources. Compared with the prior art, the method has the advantages of accurate resource allocation, high management efficiency and the like.

Description

Technical Field

[0001] The present invention relates to the field of big data technology, and particularly to a management method and system for an elderly care meal assistance platform based on big data. Background Art

[0002] The elderly care meal service has become an important part of social governance. However, the current meal assistance service model has the following problems:

[0003] 1. Low supply-demand matching efficiency: It is difficult to accurately allocate resources according to the actual needs of the elderly, resulting in unreasonable distribution of meal assistance services.

[0004] 2. High operation and management costs: In rural and remote areas, due to geographical dispersion and logistics costs, the operation sustainability is insufficient.

[0005] 3. Insufficient service coverage: The traditional service model is difficult to comprehensively cover urban and rural areas, resulting in the unmet needs of the elderly not being fully satisfied. Summary of the Invention

[0006] The purpose of the present invention is to overcome the above-mentioned defects existing in the prior art and provide a management method and system for an elderly care meal assistance platform based on big data. Through refined management and dynamic adjustment, the limitations of the traditional model are solved, the efficiency of meal assistance services is improved, resource allocation is optimized, and the high-quality development of precise and intelligent elderly care services is supported.

[0007] The purpose of the present invention can be achieved through the following technical solutions:

[0008] A management method for an elderly care meal assistance platform based on big data, the method comprising the following steps:

[0009] Collect multi-source data and store it, the multi-source data includes historical data and real-time data, wherein the historical data includes user data, order data, and meal assistance outlet data, and the real-time data includes external data, user behavior data, and logistics data;

[0010] Based on the multi-source data, use big data for meal assistance demand prediction, the meal assistance demand prediction includes geographical distribution prediction, time distribution prediction, population characteristic distribution prediction, and service type distribution prediction, and the prediction results are represented by the order volume and order type;

[0011] Obtain real-time order data, and based on the demand prediction results, use the genetic algorithm to design the shortest path and the optimal task allocation plan under limited dispatcher and vehicle resources to optimize the distribution resources.

[0012] The user data includes the user's registration information, dietary preferences, and consumption habits. The order data includes the daily order volume, order sources, order types, and order time period distribution. The meal assistance network data includes the capacity of meal assistance points, the supply of food items, inventory change records, network operation hours, and regional coverage. The external data includes weather conditions and holiday information. The user behavior data includes real-time order placement behavior and user interaction data. The logistics data includes delivery progress and road condition information.

[0013] After the multi-source data undergoes data cleaning, data format conversion, and feature extraction, it is stored based on the Hadoop HDFS distributed file system and an index is established using Elasticsearch to quickly retrieve user preferences and historical orders. Among them, before storing into the index, the multi-source data is mapped to independent index types according to different data sources. Each index type includes multiple fields for quickly locating data. The index tree is constructed based on the inverted index structure. Each index maintains an inverted list. The inverted list uses the keyword field as the primary key, and the tokenization result of the data field points to the actual storage location. A tree-like index is constructed through hierarchical nodes. In addition, a sharding mechanism is added to the index tree to divide the large-volume index into multiple sub-fragments to support parallel retrieval. A cache layer is set for high-frequency queries to cache query results. The index data is updated regularly, and compressed storage is enabled for large fields.

[0014] The geographical distribution prediction is used to predict the distribution of the meal assistance needs of the elderly in different regions. The time distribution prediction is used to predict the peak and trough periods of meal needs within a day, as well as the special fluctuations in demand during holidays. The population characteristic distribution prediction predicts the meal assistance needs based on the characteristics of the elderly. The characteristics of the elderly include age group, whether living alone, and whether disabled. The service type distribution prediction is used to predict the proportion of demand for different meal assistance service types, including dine-in demand, takeout demand, and special dietary needs. The special dietary needs include low-salt and low-sugar meals and liquid food needs.

[0015] The method further includes: based on the demand prediction results, using the genetic algorithm to optimize the meal inventory management;

[0016] The optimization of meal inventory management takes the total loss of minimizing inventory shortages and overstocking at each meal assistance point as the objective function, and takes the maximum storage capacity of the meal assistance point and the shelf-life requirements of each dish as the constraint conditions. Each chromosome is represented as a combination of the meal distribution quantities at the meal assistance network points, and the fitness function is set as the reciprocal of the inventory loss. The genetic algorithm is used for optimization to obtain the inventory management optimization plan.

[0017] The optimization of distribution resources aims to minimize the total distribution time or the total path length as the objective function, with the available time of the delivery staff, vehicle capacity, and distribution area coverage as constraints. Each chromosome is represented as a combination of the delivery staff and the path, and the fitness function is set as the reciprocal of the total distribution time. The genetic algorithm is used for optimization to obtain the optimized distribution resource plan.

[0018] The method further includes: real-time monitoring of order data and logistics data, and when the order status changes or the status of the delivery staff changes, task reallocation is performed. Among them, when a new order is received in real time, the new order is assigned to an idle delivery staff or an existing delivery task; when the status change of the delivery staff causes the task to be unable to be completed, its delivery task is transferred to other available delivery staff.

[0019] The method further includes: when the delivery path is abnormal, path planning optimization is performed, specifically including:

[0020] Avoiding traffic congestion: Adjusting the path according to the real-time road conditions, preferentially selecting unobstructed and shortest-time routes to generate a new delivery plan;

[0021] Optimizing multi-point merging: Merging adjacent delivery points into the same delivery route to reduce duplicate paths.

[0022] The method further includes: when there are changes in time constraints or weather / emergency situations, dynamic resource adjustment is performed, specifically including:

[0023] Temporarily mobilizing resources: When there is a weather / emergency situation, temporarily deploying additional delivery staff or vehicles for support;

[0024] Optimizing delivery batches: Real-time adjusting the number of delivery batches, and merging small orders into the next batch;

[0025] Adjusting task priorities: When there are changes in time constraints, updating task priorities based on the timeliness and importance of orders, and reallocating tasks using heuristic algorithms or dynamic allocation algorithms.

[0026] A management system for a pension meal assistance platform based on big data, used to implement the method described above. The system includes:

[0027] Data collection and storage module: Collecting and storing multi-source data, where the multi-source data includes historical data and real-time data. Among them, the historical data includes user data, order data, and meal assistance network data, and the real-time data includes external data, user behavior data, and logistics data;

[0028] Demand forecasting module: Based on multi-source data, big data is used to predict the demand for meal assistance. The meal assistance demand prediction includes geographical distribution prediction, time distribution prediction, population characteristic distribution prediction, and service type distribution prediction. The prediction results are represented by the order volume and order type.

[0029] Inventory optimization management module: Based on the demand forecasting results, a genetic algorithm is used to optimize the management of meal inventory.

[0030] Distribution scheduling module: Obtain real-time order data. Under the limited resources of delivery staff and vehicles, a genetic algorithm is used to design the shortest path and optimal task allocation plan for distribution resource optimization.

[0031] Service evaluation and optimization module: Optimize the service process and experience by collecting user feedback.

[0032] Compared with the prior art, the present invention has the following beneficial effects:

[0033] 1. The present invention takes into account the urban-rural regional differences and the personalized needs of different elderly people. Through demand forecasting in multiple aspects based on big data, it accurately identifies the demand for elderly meal assistance and precisely matches the meal assistance demand with the supply capacity of outlets, reducing resource waste.

[0034] 2. The present invention reduces the logistics cost by optimizing the distribution resources and reduces the management cost by optimizing the inventory management, improving the operational sustainability.

[0035] 3. Through real-time data analysis and feedback processing, the present invention realizes dynamic response and service optimization. Description of the Drawings

[0036] Figure 1 It is a flowchart of the method of the present invention. Detailed Embodiments

[0037] Next, the technical solutions in the embodiments of the present invention will be clearly and completely described in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are part of the embodiments of the present invention, rather than all the embodiments. All other embodiments obtained by those of ordinary skill in the art based on the embodiments of the present invention without creative efforts shall fall within the protection scope of the present invention.

[0038] Unless otherwise defined, the technical terms or scientific terms involved in this application shall have the ordinary meanings understood by those with ordinary skills in the technical field to which this application belongs. The words such as "a", "an", "one", "the" and the like involved in this application do not indicate a limitation in quantity and may represent a singular or plural number. The terms "include", "comprise", "have" and any variations thereof involved in this application are intended to cover non-exclusive inclusion. For example, a process, method, system, product or device that includes a series of steps or modules (units) is not limited to the listed steps or units, but may further include steps or units not listed, or may further include other steps or units inherent to these processes, methods, products or devices. The similar words such as "connected", "coupled" and "linked" involved in this application are not limited to physical or mechanical connections, but may include electrical connections, whether direct or indirect. The term "plurality" involved in this application means two or more. "And / or" describes the association relationship of associated objects and indicates that three relationships may exist. For example, "A and / or B" may represent three situations: A exists alone, A and B exist simultaneously, and B exists alone. The character " / " generally indicates that the associated objects before and after are in an "or" relationship. The terms "first", "second", "third" and the like involved in this application are only used to distinguish similar objects and do not represent a specific order for the objects.

[0039] This embodiment provides a management method for a pension meal assistance platform based on big data, as Figure 1 shown, and this method includes the following steps:

[0040] S1, collect multi-source data and store it.

[0041] The multi-source data includes historical data and real-time data. Among them, the historical data includes user data, order data, and meal assistance outlet data, and the real-time data includes external data, user behavior data, and logistics data.

[0042] (1) User data:

[0043] User information: age, gender, health status, geographical location.

[0044] Dietary preferences: low salt, low fat, low sugar, customized requirements.

[0045] Consumption habits: meal ordering frequency, consumption characteristics during holidays.

[0046] (2) Order data:

[0047] Daily order volume (decomposed by hour, day, week, month).

[0048] Order sources: elderly users, community institutions, pension institutions, etc.

[0049] Order types: dine-in, takeout, self-pickup at meal assistance points.

[0050] Order time period distribution: the time period distribution of demand for breakfast, lunch, and dinner.

[0051] (3) Meal assistance network data:

[0052] The capacity of meal assistance points, the supply of food items, records of inventory changes, the operating hours of the network points, and the regional coverage.

[0053] (4) External data:

[0054] Weather conditions: temperature, precipitation, and the impact of severe weather on demand.

[0055] Holiday information: demand fluctuations caused by legal holidays and region-specific events.

[0056] (5) User behavior data

[0057] Real-time order placement behavior: the current browsing, order placement, or order cancellation situations of users.

[0058] User interaction data: operations such as querying menus and favoriting dishes.

[0059] After the multi-source data undergoes data cleaning (handling missing values and outliers), data format conversion, and feature extraction, it is stored based on the Hadoop HDFS distributed file system and indexed using Elasticsearch to quickly retrieve user preferences and historical orders.

[0060] The specific construction structure of the index tree is mainly based on Elasticsearch or similar search engine technologies for quickly retrieving and querying user preferences, historical orders, and related meal assistance data.

[0061] The construction structure of the index tree is specifically described below.

[0062] 1. Data source layer

[0063] This is the basic data source for constructing the index tree, including the multi-source data obtained above.

[0064] 2. Data mapping layer

[0065] Before being stored in the index, all data needs to be structured and mapped to appropriate index fields.

[0066] First, each data source is mapped to an independent index type (Type), such as:

[0067] - User data index (user_index);

[0068] - Meal assistance outlet data index (merchant_index);

[0069] - Order data index (order_index);

[0070] - Logistics data index (route_index).

[0071] Each index type contains multiple fields for quickly locating data. For example:

[0072] - User data index fields:

[0073] ① user_id (unique identifier)

[0074] ② name (name)

[0075] ③ location (geographical location)

[0076] ④ health_score (health score)

[0077] ⑤ preference (dietary preferences, such as low salt, low sugar, etc.)

[0078] - Meal assistance outlet data index fields:

[0079] ① merchant_id (unique identifier of the meal assistance point)

[0080] ② menu_items (list of dishes)

[0081] ③ inventory (inventory status)

[0082] ④ ratings (user ratings)

[0083] - Order data index fields:

[0084] ① order_id (unique identifier of the order)

[0085] ② user_id (user identifier)

[0086] ③ merchant_id (identifier of the meal assistance point)

[0087] ④ delivery_status (delivery status)

[0088] ⑤ delivery_time (delivery timestamp)

[0089] - Logistics data index fields:

[0090] ① route_id (path identifier)

[0091] ②start_point (starting coordinate)

[0092] ③end_point (ending coordinate)

[0093] ④traffic_status (real-time traffic condition)

[0094] 3. Index Structure

[0095] The index tree is constructed based on the inverted index structure, and at the same time, the shard and replica mechanisms are added to improve the retrieval efficiency and data redundancy.

[0096] ①Inverted Index Structure

[0097] Each index maintains an inverted list, where: the key fields (such as user_id, order_id, etc.) serve as the primary key; the tokenized results of the data fields point to the actual storage location (i.e., the document).

[0098] Keyword field Document ID Health management doc_1,doc_3,doc_7 Geographical location: Shanghai doc_2,doc_5

[0099] ②Tree Structure

[0100] A tree-shaped index is constructed through hierarchical nodes (Node):

[0101] Root Node: major categories (such as users, meal assistance outlets, orders).

[0102] Child Node: hierarchical by attributes or fields (such as location, menu_items).

[0103] Leaf Node: points to the actual data document.

[0104] 4. Query Processing

[0105] When the user initiates a retrieval, the system quickly locates the relevant documents through the index tree:

[0106] ①Parse the query condition, such as querying for recommended meals according to user preferences.

[0107] ②Locate the root node, for example, "meal assistance outlet data".

[0108] ③Traverse the child nodes and leaf nodes to find the matching documents.

[0109] ④Return the result list and sort it according to the priority (such as score, distance, etc.).

[0110] 5. Index Optimization

[0111] To improve the retrieval efficiency and performance, the following optimizations are added to the index tree structure:

[0112] ① Sharding mechanism: Divide the index with a large amount of data into multiple sub - fragments (Shards) to support parallel retrieval.

[0113] ② Cache layer: Enable the query result cache for high - frequency queries.

[0114] ③ Automatic refresh: Regularly update the index data to ensure real - time performance.

[0115] ④ Compressed storage: Enable compressed storage for large fields (such as dish descriptions) to reduce the space occupied by the index.

[0116] The following is the process of querying meal recommendations based on user preferences and geographical location in an embodiment:

[0117] 1. Query conditions:

[0118] - User preferences: Low - salt, low - sugar

[0119] - Geographical location: A certain area in Ningbo City

[0120] 2. System process:

[0121] - Locate user_index to find user preferences and location.

[0122] - Locate merchant_index to screen merchants that match user preferences.

[0123] - Locate order_index and route_index to evaluate the delivery path and time.

[0124] 3. Return results:

[0125] - The meal list is sorted in descending order of scores.

[0126] - The optimal delivery path and estimated time.

[0127] S2. Based on multi - source data, use big data for meal assistance demand prediction. The meal assistance demand prediction includes geographical distribution prediction, time distribution prediction, population characteristic distribution prediction, and service type distribution prediction. The prediction results are represented by the order volume and order type.

[0128] Geographical distribution prediction is used to predict the distribution of meal assistance needs for the elderly in different regions, including geographical dimensions such as cities, suburban areas, and rural areas. For example, the meal assistance needs in urban areas may be concentrated near elderly apartments or community service centers, while the needs in rural areas may show a dispersed characteristic. Time distribution prediction is used to predict the peak and trough periods of meal needs within a day, as well as the special fluctuations in needs during holidays, such as the peak demand during lunch and dinner times. Population characteristic distribution prediction predicts meal assistance needs based on the characteristics of the elderly, including age group, whether living alone, and whether disabled. For example, the demand for delivery services may be higher among the elderly living alone or those with disabilities. Service type distribution prediction is used to predict the proportion of demand for different meal assistance service types, including dine-in demand, takeout demand, and special dietary needs. Special dietary needs include low-salt and low-sugar meals, liquid diet needs, etc.

[0129] In this embodiment, a demand prediction is carried out using a time series model (such as ARIMA).

[0130] (1) Principle of time series model

[0131] A time series model (such as ARIMA) is used to analyze the data trend in the time dimension and predict future values. This model consists of three core parts:

[0132] AR (Autoregressive) part: Using past order data as input variables to predict the current order volume.

[0133] I (Differencing) part: Eliminating non-stationarity in the data (such as long-term growth trends or seasonal fluctuations).

[0134] MA (Moving Average) part: Adjusting the current predicted value through historical prediction errors to optimize accuracy.

[0135] (2) Model steps

[0136] 1. Data preprocessing:

[0137] Data cleaning: Handling outliers and missing values (such as extreme peak orders or misreported order data).

[0138] Detrending and deseasonalizing: Extracting the main patterns of the time series through differencing or decomposition methods.

[0139] Data stationarity detection: Using unit root tests (such as ADF test) to verify whether the data is suitable for modeling.

[0140] 2. Model training:

[0141] Using the ARIMA model to fit historical order data to capture trends and periodic fluctuations.

[0142] Introduce exogenous variables (such as weather, holidays, etc.) to improve the model's response ability to special situations.

[0143] (3) Result generation:

[0144] Output the predicted values of the daily order volume in the future and the confidence intervals (the upper and lower limits of the predicted values).

[0145] Draw a trend chart based on the predicted values to provide an intuitive display of demand changes.

[0146] S3. Based on the demand forecasting results, use the genetic algorithm to optimize the meal inventory management;

[0147] The optimization of meal inventory management aims to minimize the total loss of inventory shortage and overstock at each meal assistance point as the objective function, with the maximum storage capacity of the meal assistance point and the shelf life requirements of each dish as the constraint conditions. Each chromosome is represented as a combination of the meal distribution quantities at the meal assistance network points, and the fitness function is set as the reciprocal of the inventory loss. The smaller the loss, the higher the fitness. Use the genetic algorithm for optimization to obtain the inventory management optimization plan.

[0148] The genetic algorithm searches for the optimal or near-optimal solution in complex resource allocation scenarios by simulating the biological evolution process (selection, crossover, and mutation). It includes the following steps:

[0149] (1) Population initialization

[0150] Randomly generate a certain number of chromosomes (initial population) to ensure that the solution space is covered as widely as possible.

[0151] (2) Genetic operations

[0152] Gradually optimize the solution by iteratively selecting the optimal chromosomes and introducing random changes:

[0153] 1. Selection: Select excellent chromosomes to enter the next generation according to the fitness values (such as the roulette wheel selection method).

[0154] 2. Crossover: Exchange some genes of two chromosomes to generate new offspring individuals (such as single-point crossover or multi-point crossover).

[0155] 3. Mutation: Randomly change some gene values of the chromosomes (such as adjusting the order of the delivery routes or the inventory distribution quantities).

[0156] (3) Termination conditions

[0157] Set the termination conditions of the genetic algorithm, such as reaching the set number of iterations or the fitness value no longer increasing significantly.

[0158] In one embodiment, the application of inventory management optimization is as follows:

[0159] Input data: Historical order data (demand forecasting results) of each meal assistance point; types of meals and inventory restrictions (such as maximum storage capacity, shelf life); current inventory status and available inventory in the distribution center.

[0160] Output result: Meal distribution table for each meal assistance point: For example, Meal Assistance Point A is allocated [10, 20, 30], and Meal Assistance Point B is allocated [15, 25, 35].

[0161] Example result: After optimizing inventory management, the inventory shortage rate of the meal assistance points is reduced from 10% to 2%, and the inventory waste rate is reduced from 8% to 3%.

[0162] S4. Obtain real-time order data. Under the limited resources of deliverymen and vehicles, use the genetic algorithm to design the shortest path and optimal task allocation plan for distribution resource optimization.

[0163] The distribution resource optimization takes minimizing the total distribution time or the total path length as the objective function, uses the available time of deliverymen, vehicle capacity, and distribution area coverage as constraints, represents each chromosome as a combination of deliverymen and paths, and sets the fitness function as the reciprocal of the total distribution time. The shorter the time, the higher the fitness. Use the genetic algorithm for optimization to obtain the distribution resource optimization plan.

[0164] The implementation process of the genetic algorithm in this step refers to that described in step S3, and will not be elaborated in this embodiment.

[0165] In one embodiment, the application of distribution resource optimization is as follows:

[0166] Input data: Delivery addresses and time requirements of meal assistance orders; availability of deliverymen and vehicles; real-time road conditions of the paths (obtained through external data such as GPS).

[0167] Output result: Deliveryman task allocation table: For example, Deliveryman A is responsible for [A→B→C→D], and Deliveryman B is responsible for [E→F→G]; optimal path planning: Allocate tasks based on the principle of the shortest time or the minimum total path length.

[0168] Example result: In a scenario with 20 delivery points and 5 deliverymen, after optimization by the genetic algorithm, the total distribution time is reduced by 20%, and the total path length is reduced by 15%.

[0169] In a preferred embodiment, the method further includes a real-time management and dynamic scheduling process. Based on Apache Kafka, a stream data processing system is built to monitor the order status and distribution progress, realize the real-time monitoring and dynamic scheduling of meal assistance services, and improve the operation efficiency. Specifically, it includes:

[0170] (1) Task reallocation

[0171] Dynamic addition of new tasks: After receiving new orders in real time, they are assigned to idle deliverymen or existing delivery tasks.

[0172] Task transfer: When a deliveryman is unable to complete a task due to personal reasons (such as illness or accident), his delivery task is promptly transferred to other available deliverymen.

[0173] (2) Route planning optimization

[0174] Avoidance of traffic congestion: Adjust the route according to real-time road conditions, and preferentially select unobstructed and shortest-time routes.

[0175] Optimization of multi-point merging: Merge adjacent delivery points into the same delivery route to reduce duplicate routes.

[0176] (3) Dynamic resource adjustment

[0177] Temporary resource mobilization: When peak periods or emergencies occur (such as a sharp increase in the number of orders during holidays), temporarily deploy additional deliverymen or vehicles for support.

[0178] Optimization of delivery batches: Adjust the number of delivery batches in real time, and merge small orders into the next batch to save resources.

[0179] The triggering conditions for this step include:

[0180] 1. Order status change: Increase or decrease in the number of orders (such as a user canceling an order).

[0181] 2. Deliveryman status change: A deliveryman is unable to complete the current task or a new deliveryman goes online.

[0182] 3. Abnormal delivery route: The original route becomes infeasible due to traffic jams, road construction, etc.

[0183] 4. Time constraint change: Adjustment of the priority of a user's order (such as the approaching specified delivery time).

[0184] 5. Weather or emergency situation: Severe weather or emergencies lead to a change in the priority of delivery tasks.

[0185] The specific adjustment process is as follows:

[0186] (1) Real-time monitoring and data update

[0187] Data collection: Use Apache Kafka to collect dynamic data such as order status, deliveryman location, and traffic conditions in real time.

[0188] Data analysis: The system analyzes the data through a rule engine to identify anomalies or optimization opportunities in the delivery process.

[0189] (2) Dynamic task allocation

[0190] Task Priority Sorting: Recalculate the priority based on the timeliness and importance of the orders.

[0191] Allocation Adjustment: Use heuristic algorithms or dynamic allocation algorithms to assign new tasks to the most suitable delivery staff or adjust the existing task assignments.

[0192] (3) Route Optimization and Scheduling

[0193] Route Replanning: Combine real-time traffic conditions to dynamically update the delivery route and generate a new delivery plan.

[0194] Resource Deployment: In high-demand scenarios, notify the backup delivery resources (such as additional delivery staff or spare vehicles) to go online for support.

[0195] (4) Exception Handling

[0196] Order-level Exceptions: Notify the delivery staff in real time for user order exceptions (such as order cancellation, change of delivery time).

[0197] Route-level Exceptions: Provide alternative routes for route interruptions (such as traffic control).

[0198] Resource-level Exceptions: When the delivery staff has equipment failures or unexpected work stoppages, quickly transfer their tasks.

[0199] In addition, in a preferred embodiment, result display and feedback are also provided. Specifically, use Tableau or ECharts to generate real-time order trend charts, delivery efficiency reports, etc.

[0200] Among them, for the real-time order trend chart, its horizontal axis (X-axis) represents the time dimension and is used to show the change trend of orders over time. Specific time granularity:

[0201] Hour: Show the order changes in each period of a day (suitable for short-term trends).

[0202] Day: Show the fluctuations in the daily order volume within a week or a month (suitable for medium-term trends).

[0203] Month: Show the changes in the total monthly order volume within a year (suitable for long-term trends).

[0204] The vertical axis (Y-axis) represents the order quantity and is used to show the total order volume or category distribution within a specific time period. Specific indicators:

[0205] Total Order Volume: The total number of all orders within a unit time.

[0206] Order Type Distribution: Such as dine-in orders, takeout orders, special needs orders (low salt, low sugar, etc.).

[0207] Regional order distribution: such as the order volume in urban areas and suburban areas.

[0208] In one implementation case, it may include:

[0209] ① Real-time order volume change chart:

[0210] Horizontal axis: hourly time points (such as "08:00, 09:00, 10:00...").

[0211] Vertical axis: total order volume per hour.

[0212] Purpose: used to monitor the order volume during peak hours and optimize resource allocation.

[0213] ② Daily order volume trend chart:

[0214] Horizontal axis: date (such as "December 1, December 2...").

[0215] Vertical axis: total daily order volume.

[0216] Purpose: to evaluate the overall demand trend of the meal assistance service in the short term.

[0217] ③ Regional order distribution map:

[0218] Horizontal axis: region (such as "Urban Area A, Suburban Area B...").

[0219] Vertical axis: total regional order volume.

[0220] Purpose: to analyze the order demand in different regions and optimize the layout of meal assistance outlets.

[0221] This embodiment also provides a management system for an elderly meal assistance platform based on big data, which is used to implement the method described above. The system includes:

[0222] Data collection and storage module: collect multi-source data and store it. The multi-source data includes historical data and real-time data. Among them, the historical data includes user data, order data, and meal assistance outlet data, and the real-time data includes external data, user behavior data, and logistics data;

[0223] Demand prediction module: based on multi-source data, use big data to predict meal assistance demand. The meal assistance demand prediction includes geographical distribution prediction, time distribution prediction, population characteristic distribution prediction, and service type distribution prediction. The prediction results are represented by order volume and order type;

[0224] Inventory optimization management module: based on the demand prediction results, use the genetic algorithm to optimize the meal inventory management;

[0225] Delivery scheduling module: Obtain real-time order data. Under the limited resources of delivery staff and vehicles, use genetic algorithms to design the shortest path and optimal task allocation plan for optimizing delivery resources.

[0226] Service evaluation and optimization module: Optimize service processes and experiences by collecting user feedback.

[0227] Those skilled in the art can clearly understand that for the convenience and brevity of description, the specific working processes of the described modules can refer to the corresponding processes in the foregoing method embodiments and will not be elaborated herein.

[0228] The above are only specific embodiments of the present invention, but the protection scope of the present invention is not limited thereto. Any person skilled in the art within the technical scope disclosed by the present invention can easily think of various equivalent modifications or substitutions, and these modifications or substitutions should be covered within the protection scope of the present invention. Therefore, the protection scope of the present invention shall be subject to the protection scope of the claims.

Claims

1. A management method for a senior care meal assistance platform based on big data, characterized in that: The method comprises the following steps: Collect and store multi-source data, including historical data and real-time data, wherein the historical data includes user data, order data, and meal assistance outlet data, and the real-time data includes external data, user behavior data, and logistics data; Based on multi-source data, big data is used to predict the demand for meal assistance, including geographical distribution prediction, time distribution prediction, population characteristic distribution prediction and service type distribution prediction. The prediction results are expressed in order quantity and order type. Obtain real-time order data and, based on demand forecast results, use genetic algorithms to design the shortest path and optimal task allocation plan to optimize delivery resources under limited delivery personnel and vehicle resources.

2. According to the big data-based elderly care meal assistance platform management method of claim 1, it is characterized in that: The user data includes the user's registration information, dietary preferences and consumption habits; the order data includes daily order volume, order source, order type and order time period distribution; the meal assistance outlet data includes meal assistance outlet capacity, meal supply, inventory change records, outlet operating hours and regional coverage; the external data includes weather conditions and holiday information; the user behavior data includes real-time ordering behavior and user interaction data; and the logistics data includes delivery progress and road condition information.

3. According to the big data-based elderly care dining assistance platform management method of claim 1, it is characterized in that: After data cleaning, data format conversion and feature extraction, the multi-source data is stored based on the Hadoop HDFS distributed file system, and indexes are established using Elasticsearch to quickly retrieve user preferences and historical orders. Before being stored in the index, the multi-source data is mapped to independent index types according to different data sources. Each index type includes multiple fields for quickly locating data. The index tree is built based on an inverted index structure. Each index maintains an inverted table. The inverted table uses a key field as a primary key. The word segmentation result of the data field points to the actual storage location. A tree index is built through hierarchical nodes. In addition, a sharding mechanism is added to the index tree to divide the index of a large amount of data into multiple sub-segments to support parallel retrieval. A cache layer is set for high-frequency queries to cache query results. The index data is updated regularly, and compressed storage is enabled for large fields.

4. According to the big data-based elderly care meal assistance platform management method of claim 1, it is characterized in that: The geographic distribution prediction is used to predict the distribution of the elderly's demand for meal assistance in different areas. The time distribution prediction is used to predict the peak and trough periods of dining demand within a day, as well as special fluctuations in demand during holidays. The population characteristic distribution prediction predicts the demand for meal assistance based on the characteristics of the elderly, including age group, whether living alone, and whether disabled. The service type distribution prediction is used to predict the demand share of different types of meal assistance services, including dine-in demand, take-out demand, and special dietary demand. The special dietary demand includes low-salt and low-sugar meals and liquid food demand.

5. According to the big data-based elderly care dining assistance platform management method of claim 1, it is characterized in that: The method further includes: optimizing food inventory management using a genetic algorithm based on the demand forecast results; The meal inventory management optimization takes minimizing the total loss of insufficient inventory and excess inventory at each meal assistance point as the objective function, takes the maximum storage capacity of the meal assistance point and the shelf life requirements of each dish as constraints, represents each chromosome as a combination of meal allocation quantities at the meal assistance point, and sets the fitness function as the inverse of inventory loss. Genetic algorithm is used for optimization to obtain an inventory management optimization plan.

6. The method for managing a big data-based elderly care dining platform according to claim 1, characterized in that: The distribution resource optimization takes minimizing the total delivery time or the total length of the route as the objective function, and takes the available time of the delivery person, the vehicle capacity, and the coverage of the distribution area as constraints. Each chromosome is represented as a combination of the delivery person and the route, and the fitness function is set to the inverse of the total delivery time. The genetic algorithm is used for optimization to obtain a distribution resource optimization solution.

7. The method for managing a big data-based elderly care and dining assistance platform according to claim 1, characterized in that: The method also includes: real-time monitoring of order data and logistics data, and reallocating tasks when the order status or the deliveryman status changes, wherein when a new order is received in real time, the new order is allocated to an idle deliveryman or an existing delivery task; when a deliveryman's status changes and he is unable to complete the task, his delivery task is transferred to other available deliverymen.

8. The method for managing a big data-based elderly care dining platform according to claim 1, characterized in that: The method further includes: when the delivery path is abnormal, performing path planning optimization, specifically including: Traffic congestion avoidance: adjust the route according to the real-time traffic conditions, give priority to the unobstructed and shortest route, and generate a new delivery plan; Multi-point merging optimization: merge adjacent delivery points into the same delivery route to reduce duplicate routes.

9. The method for managing a big data-based elderly care dining platform according to claim 1, characterized in that: The method further includes: when time constraints change or weather / emergency conditions occur, dynamically adjusting resources, specifically including: Temporary resource mobilization: When weather / emergency situations occur, temporary deployment of additional delivery personnel or vehicle support; Delivery batch optimization: adjust the delivery batch quantity in real time and merge small orders into the next batch; Task priority adjustment: When time constraints change, the task priority is updated based on the timeliness and importance of the order, and the tasks are reallocated using heuristic algorithms or dynamic allocation algorithms.

10. A big data-based elderly care and dining assistance platform management system, characterized in that: For implementing the method according to any one of claims 1 to 9, the system comprises: Data collection and storage module: collects and stores multi-source data, including historical data and real-time data. The historical data includes user data, order data, and meal assistance outlet data, and the real-time data includes external data, user behavior data, and logistics data. Demand forecasting module: Based on multi-source data, big data is used to forecast the demand for meal assistance. The forecast for meal assistance demand includes geographical distribution forecast, time distribution forecast, population characteristic distribution forecast and service type distribution forecast. The forecast results are expressed in terms of order quantity and order type. Inventory optimization management module: Based on demand forecast results, genetic algorithms are used to optimize food inventory management; Delivery scheduling module: obtains real-time order data, uses genetic algorithms to design the shortest path and optimal task allocation plan under limited delivery personnel and vehicle resources, and optimizes delivery resources; Service evaluation and optimization module: Optimize service processes and experience by collecting user feedback.

Images