Household service platform order distribution system based on big data analysis

Abstract

The invention relates to the technical field of order allocation, in particular to a housekeeping platform order allocation system based on big data analysis, which comprises the following steps of: comprehensively analyzing housekeeping service request information, identifying service types, emergency degrees of time requirements and resource requirements, and integrating to form a preliminary service order list based on clear rules; the resource allocation is ensured to have pertinence from a demand end, and the matching probability of the order and the service personnel is optimized and calculated by updating the position and state data of the service personnel in real time and adopting the ant colony algorithm, so that the order matching is quickly adjusted and accurately allocated in a dynamic scene, the low-efficiency matching possibility is reduced, the resource utilization rate is improved, and the user experience is improved. The optimal service path is designed from the angles of cost and efficiency through the greedy algorithm, the service response time is ensured to be minimized, meanwhile, the operation cost is reduced, the economical efficiency and the performability of path planning are enhanced, and the distribution efficiency, the user satisfaction degree and the overall operation efficiency of the platform are improved.

Description

Technical Field

[0001] The present invention relates to the technical field of order allocation, and particularly to a domestic service platform order allocation system based on big data analysis. Background Art

[0002] The technical field of order allocation mainly studies how to efficiently and reasonably match order requirements with available resources to achieve the optimal utilization of resources and the rapid satisfaction of user needs. The core lies in designing efficient allocation strategies and algorithms to maximize service efficiency, minimize resource waste and allocation costs.

[0003] The purpose of the domestic service platform order allocation system based on big data analysis is to comprehensively analyze order information and service resource information, design a reasonable allocation plan, ensure the fairness, efficiency and service quality improvement in the order allocation process, aiming to achieve the intelligent scheduling of domestic service resources, enhance the user experience, and improve the operation efficiency of the domestic service industry.

[0004] The existing technologies have multiple limitations in the field of order allocation. They cannot fully analyze the specific characteristics and resource requirements of order requests, and the resource allocation lacks pertinence, resulting in a long response time for urgent orders and affecting the user experience. The status and location information of service personnel cannot be updated in real time, making it difficult to support dynamic scheduling requirements, resulting in low allocation efficiency and lagging resource scheduling. Moreover, the existing technologies lack effective clustering analysis and grouping optimization methods, leading to the scattered processing of orders in close areas, increasing the redundancy and cost of path planning, and being unable to flexibly consider dynamic cost and efficiency requirements, resulting in uneconomical path selection and excessive response time. Summary of the Invention

[0005] The purpose of the present invention is to solve the drawbacks existing in the prior art, and to propose a domestic service platform order allocation system based on big data analysis.

[0006] To achieve the above purpose, the present invention adopts the following technical solutions: A domestic service platform order allocation system based on big data analysis includes:

[0007] An order receiving and screening module: Based on the domestic service request information of users, conduct data analysis on service types and expected times, identify the urgency and resource requirements, integrate relevant data, and generate a preliminary service order list;

[0008] A dynamic service personnel matching module: According to the preliminary service order list, analyze the real-time positions and statuses of service personnel, update the model through real-time data, and use the ant colony algorithm to calculate the matching probabilities between each order and available service personnel, formulate matching strategies, and generate a service matching table;

[0009] Geographical Location Optimization Module: Analyze the geographical location data of orders according to the service matching table, group orders with close locations and urgent time requirements through geographical clustering, optimize the grouping results, and generate the task set division results;

[0010] Cost-Effective Path Module: According to the task set division results, conduct path search, adopt the greedy algorithm, analyze the path data, and select the appropriate path based on cost and efficiency to generate the path optimization plan;

[0011] Scheduling Strategy Decision Module: Based on the path optimization plan, comprehensively consider the service response time, cost, and customer satisfaction, make multi-faceted balances, and determine the appropriate scheduling plan according to the service requirements and resource status to generate the scheduling decision plan;

[0012] Order Allocation Execution Module: Based on the scheduling decision plan, implement the specific matching of orders and service personnel, update the service status, ensure the correct allocation of orders, and generate the order execution status report.

[0013] As a further solution of the present invention, the order receiving and screening module includes a data analysis sub-module, a requirement integration sub-module, and a preliminary generation sub-module, where:

[0014] Data Analysis Sub-module: Based on the household service request information of users, extract and analyze the service type and expected time, perform correlation matching according to the extracted service type data and service time requirements, and generate order time characteristics and demand distribution data;

[0015] Requirement Integration Sub-module: Based on the order time characteristics and demand distribution data, perform grouping processing on the time data and service type data, analyze the correlation characteristics between the resource requirements and orders in the grouping, and conduct overall correlation integration to establish a service requirement matching information table;

[0016] Preliminary Generation Sub-module: Based on the service requirement matching information table, screen the integrated resource requirements and service types, eliminate the unexecutable order data, generate the set of orders to be executed by analyzing the resource matching conditions, and establish a preliminary service order list.

[0017] As a further solution of the present invention, the dynamic service personnel matching module includes a service personnel status analysis sub-module, a matching probability calculation sub-module, and a matching strategy generation sub-module, where:

[0018] Service Personnel Status Analysis Sub-module: Based on the preliminary service order list, analyze the current task completion situation through status data, extract the idle status and geographical location information of service personnel, classify and analyze the extracted status data, and generate a service personnel real-time status table;

[0019] Matching probability calculation sub-module: Based on the real-time status table of service personnel, combined with the service type and location data in the preliminary service order list, calculate the location distance between the geographical locations of each order and service personnel, and perform associated matching to generate a service personnel-order matching probability table;

[0020] Matching strategy generation sub-module: According to the service personnel-order matching probability table, use the ant colony algorithm to analyze the matching sorting results, determine the corresponding relationship between the priorities of service personnel and orders, formulate a matching plan in combination with the priorities and order characteristics, and generate a service matching table.

[0021] As a further solution of the present invention, the ant colony algorithm is calculated according to the formula:

[0022]

[0023] where: P ij is the matching probability value between service personnel i and order j, τ ij is the pheromone intensity of the path i→j, η ij is the heuristic factor, λ ij is the geographical distance influence factor, d ij is the geographical distance between service personnel i and order j, φ ij is the time sensitivity factor, t ij is the estimated time for service personnel i to reach order j, T is the time sensitivity threshold, ψ ij is the order demand urgency factor, α is the importance weight of the pheromone intensity, β is the importance weight of the heuristic factor, N i is the set of optional orders for service personnel i.

[0024] As a further solution of the present invention, the geographical location optimization module includes a location data extraction sub-module, an order clustering and grouping sub-module, and a grouping optimization and generation sub-module, where:

[0025] Location data extraction sub-module: Based on the service matching table, extract the geographical location information of the orders, format the extracted data and filter out invalid data, perform coordinate conversion on the valid geographical data, extract the service time requirements at the same time and divide the time intervals, and associate the converted coordinates with the divided time intervals to generate order location-time characteristic data;

[0026] Order clustering and grouping sub-module: Based on the order location-time characteristic data, perform clustering analysis on the geographical coordinate data according to spatial proximity, and classify them into preliminary groups, extract the task time requirement data of the orders within each group, perform time sorting, and generate order location-time grouping data;

[0027] Group optimization generation sub-module: Based on the order location time grouping data, calculate the cumulative time interval value and the cumulative distance value within each group according to the time interval and spatial distribution, optimize the arrangement of the location order within the group, and integrate them to generate the task set division result.

[0028] As a further aspect of the present invention, the cost-benefit path module includes a path data analysis sub-module, a path optimization calculation sub-module, and a path generation adjustment sub-module, wherein:

[0029] Path data analysis sub-module: Based on the task set division result, extract the geographical coordinate order and time interval data, analyze the geographical coordinate order to calculate the driving distance between tasks, and at the same time calculate the time interval data, analyze the time difference between tasks and determine the time interval of each task, and generate the path location time analysis result;

[0030] Path optimization calculation sub-module: Based on the path location time analysis result, use the greedy algorithm to calculate the total driving distance between tasks in the path, comprehensively calculate the time interval and the total distance, adjust the arrangement order of the path tasks and recalculate the comprehensive value, and analyze the cost and efficiency of the path through the comprehensive value to generate the path cost-benefit data;

[0031] Path generation adjustment sub-module: Based on the path cost-benefit data, adjust the task order, analyze the influence of the geographical distribution data in the path on the order, and at the same time optimize the path data by adjusting the order, and integrate them to generate the path optimization scheme.

[0032] As a further aspect of the present invention, the greedy algorithm is calculated according to the formula:

[0033]

[0034] Where: C is the comprehensive cost value of the path, s i,i+1 is the driving distance between task i and task i + 1, u i,i+1 is the average driving speed between task i and task i + 1, z i and z i+1 are the planned service time points of task i and task i + 1, Q is the maximum time interval threshold allowed between tasks, p is the time cost weight coefficient, r is the priority influence weight coefficient, v i,i+1 is the priority difference between task i and task i + 1, t is the path stability weight coefficient, w i,i+1 is the matching stability factor between task i and task i + 1.

[0035] As a further aspect of the present invention, the scheduling strategy decision module includes a service time evaluation sub-module, a cost-benefit balance sub-module, and a scheduling strategy generation sub-module, wherein:

[0036] Service time evaluation sub-module: Based on the path optimization scheme, extract the service time data of each task in the path, perform normalization processing on the time data, mark it according to the task, calculate the time difference of each task through the time-marked data, analyze the distribution characteristics of the time difference, and conduct sorting statistics to generate the task time response evaluation result;

[0037] Cost-benefit balance sub-module: Based on the task time response evaluation result, extract the time difference and the distance data of each task in the path, perform normalization processing on the time difference and the distance data, and conduct ratio calculation. At the same time, analyze the comprehensive cost of the task to generate the task cost-benefit balance data;

[0038] Scheduling strategy generation sub-module: Based on the task cost-benefit balance data, extract the priority sorting information of each task, conduct cross-analysis of the sorting information and the path optimization scheme, analyze the rationality of task allocation in the path scheme, adjust the sorting information, and combine the adjusted task order to generate the scheduling decision scheme.

[0039] As a further solution of the present invention, for analyzing the distribution characteristics of the time difference, the time difference is calculated as the difference between the start time and the end time of the task, and the analysis of the distribution characteristics of the time difference includes statistically calculating the average value, standard deviation, and distribution range of the task time difference;

[0040] For analyzing the rationality of task allocation in the path scheme, combining the adjusted task order, the rationality of path allocation is judged according to the priority, the sorting information is adjusted according to the priority to adjust the task order, and it is ensured that the path dependency relationship is not damaged. The scheduling decision scheme includes the optimized task execution order, adjustment records, and the matching degree information of the priority and the path order.

[0041] As a further solution of the present invention, the order allocation execution module includes an order matching implementation sub-module, a service status update sub-module, and an allocation status record sub-module, where:

[0042] Order matching implementation sub-module: Based on the scheduling decision scheme, extract the real-time status data of the scheduling priority queue and the service personnel, and through analyzing the relevance between the current idle status of the service personnel and the task requirements, perform matching verification on each task and the available service personnel to generate the order matching result data;

[0043] Service status update sub-module: Based on the order matching result data, extract the service personnel matching information and the task execution status data, update and mark the task status of the service personnel, associate and record the execution status of the task and the service personnel, and dynamically update the status table to generate the task execution status data;

[0044] Allocation Status Recording Sub-module: Based on the task execution status data, extract the task allocation information that has completed matching and the execution status of service personnel, mark the tasks with completed allocation as the allocated status, integrate the allocation marking data into task allocation status data, and summarize and generate an order execution status report.

[0045] Compared with the prior art, the advantages and positive effects of the present invention are as follows:

[0046] 1. In the present invention, through a comprehensive analysis of housekeeping service request information, identify the service type, the urgency of time requirements, and resource requirements, and integrate and form a preliminary service order list based on clear rules to ensure the pertinence of resource allocation from the demand side;

[0047] 2. In the present invention, by real-time updating the location and status data of service personnel, use the ant colony algorithm to optimize the calculation of the matching probability between orders and service personnel, so that order matching can be quickly adjusted and accurately allocated in a dynamic scenario, reduce the possibility of inefficient matching, and improve resource utilization rate;

[0048] 3. In the present invention, through the geographical clustering method, group the orders in the proximity area and with urgent time requirements to reduce path redundancy in the allocation process and improve the coordination and rationality of task division;

[0049] 4. In the present invention, design the optimal service path from the perspectives of cost and efficiency through the greedy algorithm, ensure the minimization of service response time, reduce operating costs at the same time, enhance the economy and executability of path planning, and improve the allocation efficiency, user satisfaction and overall operation efficiency of the platform. BRIEF DESCRIPTION OF THE DRAWINGS

[0050] Figure 1 is the system flow chart of the present invention;

[0051] Figure 2 is the process schematic diagram of the present invention;

[0052] Figure 3 is the system framework schematic diagram of the present invention. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0053] In order to make the objectives, technical solutions and advantages of the present invention clearer and more understandable, the present invention will be further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present invention and are not used to limit the present invention.

[0054] Please refer to Figure 1 and Figure 2 , the present invention provides a technical solution: A housekeeping platform order allocation system based on big data analysis includes:

[0055] Order Receiving and Screening Module: Based on the user's domestic service request information, analyze the service type and expected time for data analysis, identify the urgency and resource requirements, integrate relevant data, and generate a preliminary service order list;

[0056] Dynamic Service Personnel Matching Module: According to the preliminary service order list, analyze the real-time location and status of service personnel, update the model through real-time data, and use the ant colony algorithm to calculate the matching probability between each order and available service personnel, formulate a matching strategy, and generate a service matching table;

[0057] Geographical Location Optimization Module: According to the service matching table, analyze the geographical location data of orders, group orders with close locations and urgent time requirements through geographical clustering, optimize the grouping results, and generate the task set division result;

[0058] Cost-Effective Path Module: According to the task set division result, conduct path search, use the greedy algorithm to analyze path data, and select an appropriate path based on cost and efficiency to generate a path optimization plan;

[0059] Scheduling Strategy Decision Module: Based on the path optimization plan, comprehensively consider the service response time, cost, and customer satisfaction, conduct multi-faceted balancing, determine an appropriate scheduling plan according to service requirements and resource status, and generate a scheduling decision plan;

[0060] Order Allocation Execution Module: Based on the scheduling decision plan, implement the specific matching of orders and service personnel, update the service status, ensure the correct allocation of orders, and generate an order execution status report.

[0061] Please refer to Figure 3 , the order receiving and screening module includes a data analysis sub-module, a demand integration sub-module, and a preliminary generation sub-module, where:

[0062] Data Analysis Sub-module: Based on the user's domestic service request information, extract and analyze the service type and expected time, perform correlation matching according to the extracted service type data and service time requirements, and generate order time characteristics and demand distribution data;

[0063] Demand Integration Sub-module: Based on the order time characteristics and demand distribution data, group the time data and service type data, analyze the resource requirements and the correlation characteristics between orders in the group, and conduct overall correlation integration to establish a service demand matching information table;

[0064] Preliminary Generation Sub-module: Based on the service demand matching information table, screen the integrated resource requirements and service types, eliminate unexecutable order data, generate a set of orders to be executed by analyzing resource matching conditions, and establish a preliminary service order list;

[0065] Data analysis sub-module: Based on the household service request information of users, using the multi-attribute decision-making method, specifically the analytic hierarchy process, the service request information is divided into three attributes: service type, service time, and urgency. For the service type classification and coding, integer mapping coding is used, and different types of services are respectively mapped to integer codes from 1 to 10. For the service time requirement, it is segmented in the 24-hour format using the interval division method, and the time is divided into four intervals: 0:00 - 6:00, 6:01 - 12:00, 12:01 - 18:00, and 18:01 - 24:00, and numerically marked as 1 - 4. For the urgency, a logistic regression model is used for classification processing. The initial parameter value of the logistic regression model is 0.01, and the model parameters are updated through the maximum likelihood estimation method. During the classification processing, a data set with a batch size of 64 is used for training to generate order time characteristics and demand distribution data;

[0066] Demand integration sub-module: Based on the order time characteristics and demand distribution data, the K-means clustering algorithm is used to group the time data and service type data. The number of clustering centers is set to 4. Using the order time characteristics and service type as input features, the Euclidean distance is used to calculate the distance from each order data to the clustering center. By setting the maximum number of iterations to 300 and using dynamic update of the clustering center to complete the grouping, for each group of data, the association rule algorithm is used to calculate the association features. With the service type as the leading item and the service time as the subsequent item, the minimum support threshold is set to 0.2 and the minimum confidence threshold is set to 0.6 to extract strong association rules. The data in the grouping is matched and integrated item by item with the resource requirements to establish a service demand matching information table;

[0067] Initial generation sub-module: Based on the service demand matching information table, the integrated resource requirements and service types are screened, and the unexecutable order data is removed. The conditional screening method is used to check the resource conditions. The order data with the resource available status of 1 is included in the screening scope, and the order data that does not meet the conditions is removed. The dynamic programming algorithm is used to construct a multi-stage decision-making model. The state variables of the model are the order number and the resource number, and the decision variable is the execution result of the order-resource matching. During the dynamic programming recursion process, a discount factor of 0.95 is used and combined with the immediate reward value to calculate the matching priority. By making decisions stage by stage, all the matching data of orders and resources are accumulated to generate a set of orders to be executed and establish a preliminary service order list.

[0068] Please refer to Figure 3 The dynamic service personnel matching module includes a service personnel status analysis sub-module, a matching probability calculation sub-module, and a matching strategy generation sub-module, where:

[0069] Service staff status analysis sub-module: Based on the preliminary service order list, analyze the current task completion status through status data, extract the idle status and geographical location information of service staff, classify and analyze the extracted status data, and generate a real-time status table of service staff;

[0070] Matching probability calculation sub-module: Based on the real-time status table of service staff, combine the service type and location data in the preliminary service order list, calculate the location distance between each order and the service staff's geographical location, and perform associated matching to generate a service staff-order matching probability table;

[0071] Matching strategy generation sub-module: According to the service staff-order matching probability table, use the ant colony algorithm to analyze the matching sorting results, determine the corresponding relationship between the priorities of service staff and orders, combine the priorities and order characteristics to formulate a matching plan, and generate a service matching table;

[0072] Service staff status analysis sub-module: Based on the preliminary service order list, use the classification decision tree algorithm to analyze the task status data of service staff, extract the current task completion status and idle status data of service staff. The status data includes the task completion mark and the task completion timestamp. Use the binary classification standard to mark the task status as completed and uncompleted. Among them, the data of service staff with the status marked as completed enters the subsequent process. For the extracted geographical location information, use the reverse geocoding method to convert the longitude and latitude data into readable location data. Use the batch processing mode to perform parallel encoding on all extracted data. The encoding parameters include the longitude and latitude fields and the batch processing threshold of 1000 data items. Integrate and process the classification and geographical location information to generate a real-time status table of service staff;

[0073] Matching probability calculation sub-module: Based on the real-time status table of service staff, combine the service type and geographical location information in the preliminary service order list, use the Haversine formula to calculate the geographical location distance between each order and the service staff, convert the calculated distance data to the appropriate unit and retain two decimal places, perform associated matching on the service type and distance data, use the matching probability calculation formula to calculate the matching probability between the order and the service staff. In the formula, the distance data is used as the weight factor with a weight coefficient of 0.7, and the service type matching is used as the key factor with a weight coefficient of 0.3. Perform normalization processing on the two calculated data items to eliminate the influence of dimensions, and generate matching records one by one through the comprehensive probability calculation of associated matching to generate a service staff-order matching probability table;

[0074] Matching strategy generation sub-module: According to the service staff order matching probability table, the ant colony algorithm is used to prioritize the matching probability data. The algorithm parameters include the pheromone evaporation factor set to 0.5, the pheromone increment coefficient set to 1, the number of ants set to 50, the maximum number of iterations set to 200, and the initial pheromone value set to 0.1. The matching probability is used as heuristic information to calculate the ant path selection probability. The pheromone matrix is updated iteratively to improve the path selection accuracy. The matching scheme is gradually generated by combining the characteristic data of the order and the sorted priority information, and a service matching table is generated.

[0075] The ant colony algorithm, according to the formula:

[0076]

[0077] Where: P ij is the matching probability value between service staff i and order j, τ ij is the pheromone intensity of the path i→j, η ij is the heuristic factor, λ ij is the geographical distance influence factor, d ij is the geographical distance between service staff i and order j, φ ij is the time sensitivity factor, t ij is the estimated time for service staff i to reach order j, T is the time sensitive threshold, ψ ij is the order demand urgency factor, α is the importance weight of the pheromone intensity, β is the importance weight of the heuristic factor, N i is the set of optional orders for service staff i;

[0078] Execution process: First, initialize the path pheromone intensity τ ij , and dynamically update the pheromone value according to the historical matching records of service staff i and order j in subsequent matching, which is used to reflect the preferential selection probability of service staff for specific orders. Then calculate the heuristic factor η ij , which is determined by the product of the service staff ability coefficient and the order demand coefficient. The service staff ability coefficient is calculated based on the rating, service experience, and professional skill level on the platform. The order demand coefficient is quantified through demand dimensions such as order type and service difficulty. Then calculate the geographical distance influence factor λ ij , obtain the geographical locations of service staff and orders in real time through big data, and calculate the straight-line distance d ij between them, which reflects the influence of geographical location on the matching priority. After that, calculate the time sensitivity factor φ ij , obtain the real-time road conditions and distance estimation through big data analysis. The time sensitive threshold T is set according to the distribution law of order completion time in the platform historical data. Then calculate the order demand urgency factor ψ ij, determined by the product of the order priority level score and the ratio of the remaining completion time. The priority level is preset by the platform, and the remaining time is calculated based on the difference between the order deadline and the current time. Finally, substituting into the formula, a matching probability value P is generated through the weighted calculation of the pheromone intensity, heuristic factor, and improvement parameter ij , and sort all the orders in the optional order set N i of the service personnel i to generate a matching plan and complete the formulation of the service matching table.

[0079] Please refer to Figure 3 , the geographical location optimization module includes a location data extraction sub-module, an order clustering and grouping sub-module, and a grouping optimization and generation sub-module, where:

[0080] Location data extraction sub-module: Based on the service matching table, extract the geographical location information of the orders, format the extracted data and filter out invalid data, convert the valid geographical data into coordinates, extract the service time requirements and divide the time intervals at the same time, and associate the converted coordinates with the divided time intervals to generate order location-time characteristic data;

[0081] Order clustering and grouping sub-module: Based on the order location-time characteristic data, perform clustering analysis on the geographical coordinate data according to spatial proximity and classify them into preliminary groups, extract the task time requirement data of the orders within each group, perform time sorting, and generate order location-time grouping data;

[0082] Grouping optimization and generation sub-module: Based on the order location-time grouping data, calculate the cumulative time interval value and the cumulative distance value within each group according to the time interval and spatial distribution, optimize the arrangement of the location order within the group, and integrate them to generate the task set division result;

[0083] Location Data Extraction Sub-module: Based on the service matching table, extract the geographical location information of the order, and use the formatting processing algorithm to unify the extracted data into a standard format. The processing rules include retaining the longitude and latitude fields, restricting the range of longitude and latitude data between -180 and 180, and using anomaly detection methods to eliminate data that exceeds the range or is empty. The elimination rules include marking null values and deleting duplicate data. For valid geographical data, use the coordinate transformation method, specifically the Amap API for geographical coordinate system transformation. The transformation method parameters include the input coordinate system being WGS84 and the output coordinate system being GCJ02. The size of the batch-processed coordinate data is set to 1000 entries, and the transformation is completed item by item by calling the interface. At the same time, extract the service time requirement data and use the time formatting function to uniformly process the time data into the HH:mm format. The rule for dividing the time interval is in hours, and the intervals are set as four groups: 0-6, 6-12, 12-18, and 18-24. Associate the transformed coordinate data with the divided time intervals through the primary key field to generate order location-time characteristic data;

[0084] Order Clustering and Grouping Sub-module: Based on the order location-time characteristic data, use the clustering algorithm to perform clustering analysis on the geographical coordinate data according to spatial proximity. The clustering parameters include the radius threshold eps set to 0.01 and the minimum number of samples min_samples set to 5. Calculate the number of sample points within the neighborhood range of each data point and determine the core points, and perform clustering division according to the rule of connecting core points and density. After being classified into preliminary groups, extract the order task time data within each group, and use the quicksort algorithm to sort the time data. Set the increasing order rule during the sorting process, and achieve in-place adjustment of the data through swap sorting to generate order location-time grouping data;

[0085] Grouping Optimization and Generation Sub-module: Based on the order location-time grouping data, use the grouping optimization algorithm to calculate the time interval and spatial distribution within each group. The calculation method for the cumulative value of the time interval is to calculate the time difference between adjacent orders item by item and accumulate it. The calculation method for the cumulative value of the distance is to calculate the Haversine distance between geographical coordinate points item by item and accumulate it. Use the genetic algorithm to optimize the arrangement of the order sequence within the group. The algorithm parameters include the population size set to 50, the crossover probability set to 0.8, the mutation probability set to 0.05, and the fitness function set to the weighted sum of the time interval and the cumulative distance value, with the weight ratios being 0.6 and 0.4 respectively. Generate a new generation of population through crossover and mutation operations, and set the number of iterations to 200 times. Retain the individuals with higher fitness generation by generation during the iteration process, and perform grouping integration on the final sorting result to generate the task set division result.

[0086] Please refer to Figure 3, the cost-benefit path module includes a path data analysis sub-module, a path optimization calculation sub-module, and a path generation and adjustment sub-module, where:

[0087] Path data analysis sub-module: Based on the task set division result, extract the geographical coordinate sequence and time interval data, analyze the geographical coordinate sequence to calculate the driving distance between tasks, and at the same time calculate the time interval data, analyze the time difference between tasks and determine the time interval between tasks, and generate the path position-time analysis result;

[0088] Path optimization calculation sub-module: Based on the path position-time analysis result, use the greedy algorithm to calculate the total driving distance between tasks in the path, comprehensively calculate the time interval and the total distance, adjust the arrangement order of path tasks and recalculate the comprehensive value, and analyze the cost and efficiency of the path through the comprehensive value to generate the path cost-benefit data;

[0089] Path generation and adjustment sub-module: Based on the path cost-benefit data, adjust the task order, analyze the influence of the geographical distribution data in the path on the order, and at the same time optimize the path data by adjusting the order, and perform integration to generate the path optimization plan;

[0090] Path data analysis sub-module: Based on the task set division result, extract the geographical coordinate sequence and time interval data, use the Haversine formula to calculate the driving distance between tasks for the extracted geographical coordinate sequence, accumulate the driving distances of all task points in sequence and generate the driving distance matrix, and at the same time calculate the time difference between adjacent tasks in minutes for the time interval data, perform a traversal operation on the time difference data to eliminate invalid data less than 0, and record the time interval between tasks, integrate the driving distance and time interval data, and generate the path position-time analysis result;

[0091] Path optimization calculation sub-module: Based on the path position-time analysis result, use the greedy algorithm to calculate the total driving distance between tasks in the path, perform an iterative selection operation on the task order of each path starting from the first task point, use the minimum driving distance of the current task point as the greedy selection criterion, update the remaining task point set after selection and calculate the next driving distance, repeat this iteration until all task points are traversed, store the total driving distance data as the result variable, and at the same time perform a weighted comprehensive calculation on the time interval data and the total distance, set the time interval weight to 0.4 and the driving distance weight to 0.6, correspond the weighted value to the task order one by one and readjust the arrangement order of path tasks, recalculate the comprehensive value of the adjusted path and store it in the result matrix to generate the path cost-benefit data;

[0092] Path generation adjustment sub-module: Based on path cost-benefit data, adjust the task order, analyze the distribution impact of the order on the geographical distribution data in the path using the geographical clustering method. The geographical clustering method uses the agglomerative clustering algorithm based on hierarchical clustering. Set the initial number of clusters as the total number of task points and gradually merge the nearest neighbor clusters until the set cluster number limit is reached. Take the task order in the clustering result as the adjustment priority order, optimize the path data through the dynamic programming algorithm for the adjusted order. The state variables of the dynamic programming model are the index of the current task point and the set of completed task points. Calculate the cumulative distance and time consumption of the path in the recurrence formula and retain the record of the optimal path. Integrate the optimized path data and generate a path optimization plan.

[0093] The greedy algorithm, according to the formula:

[0094]

[0095] where: C is the comprehensive cost value of the path, s i,i+1 is the driving distance between task i and task i + 1, u i,i+1 is the average driving speed between task i and task i + 1, z i and z i+1 are the planned service time points of task i and task i + 1, Q is the maximum time interval threshold allowed between tasks, p is the time cost weight coefficient, r is the priority impact weight coefficient, v i,i+1 is the priority difference between task i and task i + 1, t is the path stability weight coefficient, w i,i+1 is the matching stability factor between task i and task i + 1;

[0096] Execution process: First, obtain the driving distance s i,i+1 between tasks in real time, calculate the shortest path distance between task i and task i + 1 through the big data map service. Then calculate the average driving speed u i,i+1 , obtain the traffic conditions of the current road section through real-time traffic data, and determine the driving speed of the service personnel between tasks, which is used to measure the impact of path time consumption on the cost. Then calculate the time interval term The time cost weight coefficient p is determined by analyzing the impact of task delay on customer satisfaction and platform economic benefits in historical orders. The time interval z i+1 -z i is calculated from the difference in planned service time points, and the allowed time interval threshold Q is set through statistical analysis of the time intervals between tasks in historical data. Then calculate the priority difference term r·v i,i+1 , the priority weight coefficient r is determined by analyzing the impact of task priority on the path optimization goal, and the priority difference v i,i+1is the difference in importance weights set for task i and task i+1 in the platform, including the difference between urgent tasks and ordinary tasks. Finally, the matching stability term t·w is calculated i,i+1 , the path stability weight coefficient t is determined by analyzing the impact of the service time fluctuation of historical orders on the flexibility of path adjustment, and the matching stability factor w i,i+1 Specifically, it is the product of the service time volatility between tasks and the current time difference, used to measure the impact of the time change between tasks on path stability. After integrating all parameters, the arrangement order of tasks is dynamically adjusted to generate the comprehensive cost value C of the path, and based on the comprehensive cost value, the cost and efficiency of the path are analyzed to optimize the task allocation plan and generate path cost-benefit data.

[0097] Please refer to Figure 3 , the scheduling policy decision module includes a service time evaluation sub-module, a cost-benefit balance sub-module, and a scheduling policy generation sub-module, where:[[]]

[0098] Service time evaluation sub-module: Based on the path optimization plan, extract the service time data of each task in the path, normalize the time data and mark it by task, calculate the time difference of each task through the time-marked data, analyze the distribution characteristics of the time difference and perform sorting statistics to generate the task time response evaluation result;

[0099] Cost-benefit balance sub-module: Based on the task time response evaluation result, extract the time difference and the distance data of each task in the path, normalize the time difference and distance data, and perform ratio calculation. At the same time, analyze the comprehensive cost of the task to generate task cost-benefit balance data;

[0100] Scheduling policy generation sub-module: Based on the task cost-benefit balance data, extract the priority sorting information of each task, cross-analyze the sorting information with the path optimization plan, analyze the rationality of task allocation in the path plan and adjust the sorting information, and combine the adjusted task order to generate a scheduling decision plan;

[0101] Service time evaluation sub-module: Based on the path optimization scheme, extract the service time data of each task in the path, and use the time formatting method to normalize the time data. The formatting rule is to uniformly convert the time data to the 24-hour system and store it in the HH:mm:ss format. Perform an association operation on the normalized time data according to the task marking field, correspond the time data with the task number and form a marking table. Use the time marking table to calculate the time difference of each task. The time difference calculation method is to calculate and record the absolute value of the service time difference between adjacent tasks. Perform statistical analysis on all time difference data and use the histogram method to draw the difference distribution diagram. The difference interval is set to 5 minutes per group and the frequency of each group is counted. Perform a sorting operation on the distribution characteristics and sort them from high to low according to the frequency to generate a statistical table, and generate the task time response evaluation result;

[0102] Cost-benefit balance sub-module: Based on the task time response evaluation result, extract the time difference and the distance data of each task in the path, and use the normalization method to process the time difference and distance data. The normalization rule is to map the data to between 0 and 1. The normalization function is that the maximum and minimum values in the normalization formula are taken from the time difference and distance data sets. Perform a ratio calculation on the normalized data. The calculation rule is that the time difference ratio is set to 0.4 and the distance ratio is set to 0.6. Use the weighted sum method to accumulate and calculate the ratio results to obtain the comprehensive cost data of each task. Use the piecewise linear regression method to analyze the corresponding relationship between the comprehensive cost data and the task number. The number of breakpoints is set to 5 and the linear model is fitted respectively to estimate the cost change trend. Organize the regression results into a table and generate the task cost-benefit balance data;

[0103] Scheduling strategy generation sub-module: Based on the task cost-benefit balance data, extract the priority sorting information of each task, and use the task priority algorithm, specifically the TOPSIS method, to adjust the priority sorting. The TOPSIS method calculation parameters include the calculation of the standardized distance between the positive ideal solution and the negative ideal solution. The positive ideal solution is the minimum comprehensive cost corresponding to the task with the highest priority, and the negative ideal solution is the maximum comprehensive cost corresponding to the task with the lowest priority. The standardized distance calculation rule is the Euclidean distance between the comprehensive cost of each task and the ideal solution. Sort the calculated results and record the corresponding order of the tasks. Perform a cross-analysis on the adjusted sorting information and the path optimization scheme. The cross-analysis method is the correlation test between the comprehensive cost of the path task and the priority of the sorted task. The test index is calculated using the Pearson correlation coefficient. Set the correlation coefficient threshold to 0.8 to screen the data with higher correlation and integrate the adjusted task order to generate a scheduling decision plan.

[0104] Analyze the distribution characteristics of the time difference, where the time difference is calculated as the difference between the start time and the end time of a task. The analysis of the distribution characteristics of the time difference includes statistical analysis of the average value, standard deviation, and distribution range of the task time difference;

[0105] Analyze the rationality of task allocation in the path plan. Combining with the adjusted task order, the rationality of path allocation is judged according to the priority. The adjusted sorting information adjusts the task order according to the priority and ensures that the path dependency relationship is not damaged. The scheduling decision plan includes the optimized task execution order, adjustment records, and the matching degree information of the priority and path order.

[0106] Please refer to Figure 3 , The order allocation execution module includes an order matching implementation sub-module, a service status update sub-module, and an allocation status record sub-module, where:

[0107] Order matching implementation sub-module: Based on the scheduling decision plan, extract the real-time status data of the scheduling priority queue and service personnel. By analyzing the relevance between the current idle status of service personnel and task requirements, match and verify each task with available service personnel to generate order matching result data;

[0108] Service status update sub-module: Based on the order matching result data, extract the service personnel matching information and task execution status data, update and mark the task status of service personnel, associate and record the execution status of tasks and service personnel, and dynamically update the status table to generate task execution status data;

[0109] Allocation status record sub-module: Based on the task execution status data, extract the task allocation information and service personnel execution status that have been successfully matched, mark the allocated tasks as the allocated status, integrate the allocation mark data into task allocation status data, and summarize and generate an order execution status report;

[0110] Order matching implementation sub-module: Based on the scheduling decision plan, extract the real-time status data of the scheduling priority queue and service personnel. Use the bipartite graph matching algorithm to verify the matching of tasks and service personnel. In the bipartite graph, set tasks as the left vertex set, set service personnel as the right vertex set, and the weight is the matching priority between service personnel and tasks. Perform maximum weight matching on the bipartite graph. In the initialization stage of the algorithm, construct an adjacency matrix, where the elements in the adjacency matrix are matching priority values. In the search stage, find unmatched vertices in the path through depth-first search, mark the path as a matching path, dynamically add unmatched tasks and service personnel to the path, and update the matching status. Output the matching result and store it in the result matrix to generate order matching result data;

[0111] Service Status Update Sub-module: Based on the order matching result data, extract the service personnel matching information and task execution status data, and use the status conversion function to dynamically update the task status of service personnel. The parameters of the status conversion function include task number, service personnel number, and task status flag value. The initial status flag value of 0 indicates idle, and it is updated to 1 during task assignment to indicate in execution. Use the service personnel number as the primary key field to associate and record the task number and service personnel status, and perform dynamic update on the status data in batch mode. The number of status update records for each batch process is set to 500. After the update is completed, dynamically insert the task number and status flag into the status table to generate task execution status data;

[0112] Allocation Status Record Sub-module: Based on the task execution status data, extract the task allocation information and service personnel execution status that have completed matching, and use the hash table storage method to quickly query and store the task allocation information and status data. The input of the hash function is the task number and service personnel number, and the output is the task allocation flag value. The allocated task flag value is 2, and the unallocated task flag value is 0. Integrate the task allocation flag value with the service personnel status value in the hash table, and store the integrated task allocation status data in the task allocation status table in the form of key-value pairs. Use the allocation status table to batch summarize the data and generate an order execution status report.

[0113] The above is only a preferred embodiment of the present invention, and it does not limit the present invention in other forms. Any person skilled in the art may use the disclosed technical content to make changes or modifications into equivalent embodiments with equivalent changes and apply them to other fields. However, as long as it does not depart from the technical solution content of the present invention, any simple modification, equivalent change, and modification made to the above embodiments based on the technical essence of the present invention still fall within the protection scope of the technical solution of the present invention.

Claims

1. A housekeeping platform order distribution system based on big data analysis, characterized in that: The system comprises: Order receiving and screening module: Based on the user's housekeeping service request information, it conducts data analysis on the service type and expected time, identifies the urgency and resource requirements, integrates relevant data, and generates a preliminary service order list; Dynamic service personnel matching module: based on the preliminary service order list, analyzes the real-time location and status of service personnel, updates the model through real-time data, and uses ant colony algorithm to calculate the matching probability between each order and available service personnel, formulates matching strategies, and generates a service matching table; Geographic location optimization module: Analyze the geographic location data of the order according to the service matching table, group the orders with close locations and urgent time requirements through geographic clustering, optimize the grouping results, and generate task set division results; Cost-effective path module: performs path search based on the task set division results, uses a greedy algorithm to analyze path data, selects an adaptive path based on cost and efficiency, and generates a path optimization solution; Scheduling strategy decision module: Based on the path optimization plan, the module comprehensively considers service response time, cost and customer satisfaction, makes a multi-faceted balance, determines the adaptive scheduling plan according to service demand and resource status, and generates a scheduling decision plan; Order allocation and execution module: Based on the scheduling decision plan, implement specific order and service personnel matching, update service status, ensure correct order allocation, and generate order execution status report.

2. The housekeeping platform order distribution system based on big data analysis according to claim 1 is characterized in that: The order receiving and screening module includes a data analysis submodule, a demand integration submodule and a preliminary generation submodule, wherein: Data analysis submodule: Based on the user's housekeeping service request information, extract and analyze the service type and expected time, associate and match the extracted service type data with the service time requirements, and generate order time characteristics and demand distribution data; Demand integration submodule: based on the order time characteristics and demand distribution data, the time data and service type data are grouped and processed, the correlation characteristics between the resource demand and the order in the group are analyzed, and the overall correlation integration is performed to establish a service demand matching information table; Initial generation submodule: Based on the service demand matching information table, the integrated resource requirements and service types are screened, unexecutable order data is eliminated, a set of orders to be executed is generated by analyzing resource matching conditions, and a preliminary service order list is established.

3. The housekeeping platform order distribution system based on big data analysis according to claim 1 is characterized in that: The dynamic service personnel matching module includes a service personnel status analysis submodule, a matching probability calculation submodule and a matching strategy generation submodule, wherein: Service personnel status analysis submodule: Based on the preliminary service order list, analyze the current task completion status through status data, extract the idle status and geographical location information of the service personnel, classify and analyze the extracted status data, and generate a real-time status table of the service personnel; Matching probability calculation submodule: Based on the service personnel real-time status table, combined with the service type and location data in the preliminary service order list, the location distance between each order and the service personnel's geographical location is calculated, and associated matching is performed to generate a service personnel order matching probability table; Matching strategy generation submodule: According to the service personnel order matching probability table, the ant colony algorithm is used to analyze the matching sorting results, determine the priority correspondence between service personnel and orders, formulate a matching plan based on the priority and order characteristics, and generate a service matching table.

4. The housekeeping platform order distribution system based on big data analysis according to claim 3 is characterized in that: The ant colony algorithm is based on the formula: Where: P ij is the matching probability value between service personnel i and order j, τ ij is the pheromone intensity of path i→j, η ij is the inspiration factor, λ ij is the geographical distance factor, d ij is the geographical distance between service personnel i and order j, φ ij is the time sensitivity factor, t ij is the estimated time for service personnel i to arrive at order j, T is the time-sensitive threshold, ψ ij is the order demand urgency factor, α is the importance weight of pheromone intensity, β is the importance weight of the heuristic factor, N i A collection of optional orders for agent i.

5. The housekeeping platform order distribution system based on big data analysis according to claim 1 is characterized in that: The geographic location optimization module includes a location data extraction submodule, an order clustering grouping submodule and a grouping optimization generation submodule, wherein: Location data extraction submodule: based on the service matching table, extract the geographic location information of the order, format the extracted data and filter out invalid data, convert the valid geographic data into coordinates, extract the service time requirements and divide the time intervals, and associate the converted coordinates with the divided time intervals to generate order location time characteristic data; Order clustering and grouping submodule: Based on the order location time characteristic data, the geographic coordinate data is clustered and analyzed according to spatial proximity, and classified into preliminary groups, the task time requirement data of the orders in each group is extracted, and the time is sorted to generate order location time grouping data; Group optimization generation submodule: Based on the order location time grouping data, the cumulative time interval value and the cumulative distance value in each group are calculated according to the time interval and spatial distribution, the position order in the group is optimized and arranged, and integrated to generate the task set division result.

6. The housekeeping platform order distribution system based on big data analysis according to claim 1 is characterized in that: The cost-effective path module includes a path data analysis submodule, a path optimization calculation submodule and a path generation adjustment submodule, wherein: Path data analysis submodule: based on the task set division result, extract the geographic coordinate sequence and time interval data, analyze the geographic coordinate sequence to calculate the driving distance between each task, and calculate the time interval data at the same time, analyze the time difference between tasks and determine the time interval of each task, and generate the path location time analysis result; Path optimization calculation submodule: Based on the path position time analysis results, a greedy algorithm is used to calculate the total driving distance between tasks in the path, the time interval and the total distance are comprehensively calculated, the arrangement order of the path tasks is adjusted and the comprehensive value is recalculated, the cost and efficiency of the path are analyzed through the comprehensive value, and the path cost-effectiveness data is generated; Path generation and adjustment submodule: Based on the path cost-effectiveness data, the task sequence is adjusted, the effect of the geographical distribution data in the path on the sequence is analyzed, and the path data is optimized by adjusting the sequence, and integrated to generate a path optimization plan.

7. The housekeeping platform order distribution system based on big data analysis according to claim 6 is characterized in that: The greedy algorithm, according to the formula: Where: C is the comprehensive cost value of the path, s i,i+1 is the driving distance between task i and task i+1, u i,i+1 is the average driving speed between task i and task i+1, z i and z i+1 is the planned service time point of task i and task i+1, Q is the maximum time interval threshold allowed between tasks, p is the time cost weight coefficient, r is the priority impact weight coefficient, and v i,i+1 is the priority difference between task i and task i+1, t is the path stability weight coefficient, w i,i+1 is the matching stability factor between task i and task i+1.

8. The housekeeping platform order distribution system based on big data analysis according to claim 1 is characterized in that: The scheduling strategy decision module includes a service time evaluation submodule, a cost-benefit balance submodule and a scheduling strategy generation submodule, wherein: Service time evaluation submodule: based on the path optimization scheme, extract the service time data of each task in the path, organize the time data and mark them by task, calculate the time difference of each task through the time mark data, analyze the distribution characteristics of the time difference and perform sorting statistics, and generate the task time response evaluation result; Cost-benefit balance submodule: based on the task time response evaluation results, extract the time difference and the distance data of each task in the path, normalize the time difference and distance data, perform proportional calculation, analyze the comprehensive cost of the task, and generate task cost-benefit balance data; Scheduling strategy generation submodule: Based on the task cost-benefit balance data, extract the priority sorting information of each task, cross-analyze the sorting information with the path optimization plan, analyze the rationality of task allocation in the path plan and adjust the sorting information, and generate a scheduling decision plan based on the adjusted task sequence.

9. The housekeeping platform order distribution system based on big data analysis according to claim 8 is characterized in that: The time difference is calculated as the difference between the start time and the end time of the task, and the time difference distribution characteristic analysis includes statistically analyzing the mean value, standard deviation and distribution range of the task time difference; The rationality of task allocation in the analysis path plan is combined with the adjusted task sequence. The rationality of path allocation is judged according to priority. The sorting information is adjusted to adjust the task sequence according to the priority, and it is ensured that the path dependency is not destroyed. The scheduling decision plan includes the optimized task execution sequence, adjustment records, and matching information of priority and path sequence.

10. The housekeeping platform order distribution system based on big data analysis according to claim 1 is characterized in that: The order allocation execution module includes an order matching implementation submodule, a service status update submodule and an allocation status recording submodule, wherein: Order matching implementation submodule: Based on the scheduling decision plan, extract the real-time status data of the scheduling priority queue and the service personnel, analyze the correlation between the current idle status of the service personnel and the task requirements, match and verify each task with the available service personnel, and generate order matching result data; Service status update submodule: based on the order matching result data, extract service personnel matching information and task execution status data, update the task status of the service personnel, associate the task with the service personnel's execution status and dynamically update the status table to generate task execution status data; Assignment status recording submodule: Based on the task execution status data, extract the completed matching task assignment information and the execution status of the service personnel, mark the assigned tasks as assigned, integrate the assignment mark data into task assignment status data and summarize and generate an order execution status report.

Images