Property management work order acceptance service system

Abstract

The invention discloses a property management work order acceptance service system, and relates to the technical field of property management, and the system comprises an acquisition and preprocessing module which collects user behavior data and carries out the preprocessing of the user behavior data; the prediction work order generation module is used for predicting and generating a prediction work order by constructing a decision tree model based on the preprocessed user behavior data; and the allocation result generation module is used for processing the predicted work order by utilizing big data analysis based on the predicted work order to obtain a priority sequence of the work order processed by the maintenance personnel, and carrying out resource scheduling, optimizing the allocation of the predicted work order and generating a work order allocation result through a particle swarm optimization algorithm. According to the method, the decision tree model is constructed to predict the preprocessed user behavior data and generate the predicted work order, so that the accuracy of work order generation is improved, manual judgment errors are reduced, potential problems are predicted in advance, and the problems of response delay and non-uniform processing are avoided.

Description

Technical Field

[0001] The present invention relates to the technical field of property management, and in particular to a property management work order acceptance service system. Background Art

[0002] With the rapid development of the Internet of Things (IoT) technology, the property management industry is gradually moving towards intelligent and digital development. Traditionally, the process for accepting work orders in property management relied primarily on manual processing, resulting in low efficiency in the generation, assignment, and execution of work orders and their potential for human interference. In recent years, with the rise of big data analytics and artificial intelligence (AI) technologies, intelligent property management systems have been gradually adopted. In these systems, data acquisition modules collect real-time data on equipment and user behavior. By analyzing this data, early warnings of equipment failures can be provided, and corresponding repair work orders can be generated. At the same time, AI-based work order prediction and resource scheduling algorithms are becoming increasingly important tools for improving work order management efficiency and service quality. In recent years, the application of intelligent algorithms such as decision tree algorithms and particle swarm optimization algorithms in work order prediction and assignment has significantly improved the automation level of property management and the timeliness of work order processing.

[0003] However, the existing property management work order acceptance service system still has some shortcomings, especially in the accuracy of work order prediction and the efficiency of resource scheduling. First, although existing technologies predict work orders by analyzing historical data, due to the lack of sufficient user behavior data and real-time monitoring data, existing systems are often unable to accurately predict user needs, resulting in insufficient timeliness and accuracy in work order generation, and delayed response problems are relatively common. Second, existing work order resource scheduling methods mainly rely on simple priority sorting, failing to fully consider multi-dimensional factors such as maintenance personnel's skills, tools, and geographical location, resulting in the rationality and efficiency of work order allocation needing to be improved. Summary of the Invention

[0004] In view of the above existing problems, the present invention is proposed.

[0005] Therefore, the present invention provides a property management work order acceptance service system to solve the problems of uneven work order distribution and low resource utilization.

[0006] In order to solve the above technical problems, the present invention provides the following technical solutions:

[0007] The present invention provides a property management work order acceptance service system, which includes: a collection and preprocessing module, which collects user behavior data and preprocesses the user behavior data; a predicted work order generation module, which makes predictions based on the preprocessed user behavior data by building a decision tree model to generate predicted work orders; an allocation result generation module, which processes the predicted work orders using big data analysis based on the predicted work orders to obtain priority rankings for maintenance personnel to handle the work orders, and performs resource scheduling through a particle swarm optimization algorithm to optimize the allocation of predicted work orders and generate work order allocation results; a maintenance report generation module, which enables maintenance personnel to dynamically adjust the work order execution progress through mobile terminal devices based on the work order allocation results and generate maintenance reports; and a scoring and feedback module, which pushes work order completion notifications to users based on maintenance completion to obtain user scoring and feedback data.

[0008] As a preferred solution of the property management work order acceptance service system described in the present invention, the user behavior data includes equipment usage frequency, historical fault records, current, voltage, temperature and humidity, and user repair report records.

[0009] As a preferred solution of the property management work order acceptance service system of the present invention, the user behavior data is pre-processed, and the specific steps are as follows:

[0010] Use Gaussian filter to filter noise from collected user behavior data;

[0011] Detect and delete duplicate records through deduplication algorithms;

[0012] Use correlation analysis and mutual information methods to filter features related to fault occurrence and work order priority;

[0013] The data scale is unified by standardizing the numerical features.

[0014] As a preferred solution of the property management work order acceptance service system of the present invention, wherein: based on the pre-processed user behavior data, a decision tree model is constructed to make predictions and generate predicted work orders. The specific steps are as follows:

[0015] Select the CART algorithm according to the nature of the task and set the decision tree model parameters;

[0016] Using the cleaned data set, the optimal features are selected for splitting based on the Gini index to construct a decision tree;

[0017] Use cross-validation to evaluate the decision tree model, adjust hyperparameters and perform pruning to optimize the decision tree model;

[0018] Apply the trained decision tree model to user behavior data, output the prediction results, and generate a decision tree model;

[0019] Input the pre-processed user behavior data into the decision tree model, make predictions, and generate prediction work orders.

[0020] As a preferred solution of the property management work order acceptance service system of the present invention, wherein: based on the predicted work order, the predicted work order is processed using big data analysis to obtain the priority ranking of the maintenance personnel processing the work order. The specific steps are as follows:

[0021] Based on the predicted work order, collect the data of the predicted work order and the information of the maintenance personnel;

[0022] Quantify and comprehensively sort the predicted work order data, and classify the predicted work orders into three levels of urgency;

[0023] Based on the maintenance personnel's skills and the predicted fault type of the work order, a scoring algorithm is used to score and obtain the maintenance personnel's maintenance skill priority;

[0024] Based on the maintenance personnel's maintenance tools and predicted work order maintenance needs, a scoring algorithm is used to score and obtain the priority of the maintenance personnel's available tools;

[0025] Based on the location information of the maintenance personnel and the predicted work order location, the Euclidean distance between the maintenance personnel and the predicted work order location is calculated and evaluated to obtain the distance priority of the maintenance personnel;

[0026] A comprehensive evaluation is performed on the maintenance personnel's maintenance skill priority, the maintenance personnel's available tool priority, and the maintenance personnel's distance priority to obtain the maintenance personnel's priority;

[0027] The predicted urgency of the work order and the priority of the maintenance personnel are comprehensively evaluated and ranked to obtain the priority of the maintenance personnel in handling the work order.

[0028] As a preferred solution of the property management work order acceptance service system of the present invention, wherein: the resource scheduling is performed through the particle swarm optimization algorithm, the work order allocation is optimized and predicted, and the work order allocation result is generated. The specific steps are as follows:

[0029] The initial state of the particle swarm is set according to the maintenance personnel's work order processing priority, and the initial work order allocation plan is randomly generated;

[0030] Based on the initial work order allocation plan, the quality of the work order allocation plan is evaluated through the fitness function;

[0031] Update the particle solution based on the quality of the work order allocation solution and the feedback of the fitness function;

[0032] Based on the solution of the updated particles, the global optimal solution is selected as the final allocation solution through the fitness function;

[0033] Based on the priority of the maintenance personnel in handling the work order, it is assigned to the maintenance personnel and the work order assignment result is generated.

[0034] As a preferred solution of the property management work order acceptance service system of the present invention, wherein: based on the solution of the updated particles, the global optimal solution is selected as the final allocation solution through the fitness function. The specific steps are as follows:

[0035] Based on the solution of the updated particles, the work order allocation plan is input into the fitness function for evaluation to obtain the fitness value of the work order allocation plan;

[0036] Use roulette wheel selection to select the solution with the best fitness value from the current solution set;

[0037] The solution with the best fitness value is selected as the parent, and crossover and mutation operations are performed to generate a new work order allocation solution.

[0038] Input the new work order allocation plan into the fitness function and calculate its fitness value;

[0039] Compare the new fitness value with the old fitness value and select the solution with higher fitness as the global optimal solution;

[0040] The work order allocation plan with the global optimal solution is used as the final allocation plan.

[0041] As a preferred solution of the property management work order acceptance service system of the present invention, wherein: according to the work order allocation result, the maintenance personnel dynamically adjust the work order execution progress and generate a maintenance report through the mobile terminal device. The specific steps are as follows:

[0042] Based on the work order assignment results, maintenance personnel receive the task and use mobile devices to update the maintenance progress in real time;

[0043] Predict work order completion time and progress based on real-time uploaded data, use the Kalman filter algorithm to dynamically adjust the work order execution progress and address any delays;

[0044] When the maintenance progress does not match the expected time, the Kalman filter algorithm will automatically adjust the expected completion time and perform resource scheduling;

[0045] The maintenance personnel completes the work order task and confirms the maintenance status as completed, and a maintenance report is automatically generated.

[0046] As a preferred solution of the property management work order acceptance service system of the present invention, when the maintenance progress does not match the expected time, the Kalman filter algorithm will automatically adjust the expected completion time and perform resource scheduling. The specific steps are as follows:

[0047] Track the completion of tasks in real time based on the initial estimated time of the work order and the actual uploaded progress data;

[0048] By regularly calculating the actual completion time of the current maintenance step and comparing it with the set estimated time, the probability of delay is assessed;

[0049] According to the delay, the variance of the measurement error and the process error is determined to obtain the Kalman gain;

[0050] Update the estimated completion time of the work order task based on the deviation between the Kalman gain and the actual completion time;

[0051] Automatically find other idle maintenance personnel to complete unfinished work order tasks through the scheduling algorithm.

[0052] As a preferred solution of the property management work order acceptance service system of the present invention, wherein: based on the maintenance report, the work order completion notification is pushed to the user, and the user rating and feedback data are obtained. The specific steps are as follows:

[0053] Automatically generate a work order completion push report based on the maintenance report, and push the work order completion to the user according to the notification receiving method pre-selected by the user;

[0054] And push a link to the user feedback form, where users can fill in service ratings, maintenance personnel performance, problem resolution status, and suggestions;

[0055] The collected user feedback data is stored in the database, and user ratings and feedback data are obtained through analysis.

[0056] The beneficial effects of the present invention are as follows: by constructing a decision tree model to predict preprocessed user behavior data and generate predicted work orders, the accuracy of work order generation is improved, manual judgment errors are reduced, potential problems are foreseen in advance, and response delays and uneven processing are avoided. Subsequently, based on the predicted work orders, a particle swarm optimization algorithm is used to schedule and allocate resources for the work orders, taking into account factors such as maintenance personnel skills, tools, and distance to ensure optimal work order allocation. This optimization algorithm improves resource scheduling efficiency, reduces work order delays, optimizes the workload of maintenance personnel, and significantly improves the efficiency and quality of work order processing, thereby improving the overall effectiveness of property management services. BRIEF DESCRIPTION OF THE DRAWINGS

[0057] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the following briefly introduces the drawings required for use in the description of the embodiments. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative work.

[0058] Figure 1 This is a schematic diagram of the property management work order acceptance service system in Example 1.

[0059] Figure 2 This is a schematic diagram of the predicted work order generation in Example 1. DETAILED DESCRIPTION

[0060] In order to make the above-mentioned objects, features and advantages of the present invention more obvious and easy to understand, the specific embodiments of the present invention are described in detail below with reference to the accompanying drawings.

[0061] In the following description, many specific details are set forth to facilitate a full understanding of the present invention. However, the present invention may also be implemented in other ways different from those described herein. Those skilled in the art may make similar generalizations without violating the connotation of the present invention. Therefore, the present invention is not limited to the specific embodiments disclosed below.

[0062] Secondly, the term "one embodiment" or "embodiment" herein refers to a specific feature, structure, or characteristic that may be included in at least one implementation of the present invention. The phrase "in one embodiment" appearing in various places throughout this specification does not necessarily refer to the same embodiment, nor does it refer to a separate or selective embodiment that is mutually exclusive of other embodiments.

[0063] Example 1, with reference to Figure 1 and Figure 2 , which is the first embodiment of the present invention, provides a property management work order acceptance service system, including the following steps:

[0064] The collection and preprocessing module collects user behavior data and preprocesses the user behavior data.

[0065] User behavior data includes device usage frequency, historical fault records, current, voltage, temperature and humidity, and user repair records;

[0066] It should be noted that equipment usage frequency refers to the number of times a device is used or the operating time within a certain period of time. Historical fault records record past equipment faults, including fault type, occurrence time, and repair history. Current refers to the current data drawn by the device during operation. Voltage refers to the voltage value drawn during operation. Temperature and humidity refer to the temperature and humidity of the device or environment. User repair report records record relevant information about user repair reports, such as a description of the fault, the time and location of the report.

[0067] Use Gaussian filter to filter noise from collected user behavior data;

[0068] It should be noted that the Gaussian filter is a common digital signal processing tool, widely used for signal smoothing and denoising. Its main purpose is to remove random noise from data, improve data quality, and ensure the accuracy of subsequent analysis and prediction.

[0069] Detect and delete duplicate records through deduplication algorithms;

[0070] It should be noted that duplicate or redundant records should be deleted to improve the quality of the data and ensure the accuracy of subsequent analysis and processing.

[0071] Use correlation analysis and mutual information methods to filter features related to fault occurrence and work order priority;

[0072] It should be noted that by using the mutual information method to select features with high information content, the system can more comprehensively understand the relationship between various factors and the target variable, thereby better capturing potential and complex correlations during the feature selection process. This helps improve the accuracy of fault prediction and work order prioritization.

[0073] The data scale is unified by standardizing the numerical features.

[0074] It should be noted that after normalization, features with different numerical ranges will be on the same scale, so that all features have equal influence in the machine learning model, avoiding the disproportionate impact of certain features on the results due to their large value range.

[0075] The prediction work order generation module builds a decision tree model based on the preprocessed user behavior data to make predictions and generate prediction work orders.

[0076] Select the CART algorithm according to the nature of the task and set the decision tree model parameters;

[0077] It should be noted that setting the decision tree model parameters, maximum depth, and setting the maximum depth of the decision tree can control the complexity of the tree. Minimum sample split number, this parameter specifies the minimum number of samples required for a node to split. Minimum leaf node sample number, this parameter controls how many samples the leaf nodes of the tree contain at least. Maximum number of features, this parameter sets the maximum number of features considered for each split. Gini index, the CART algorithm selects the best splitting feature by calculating the Gini index of each node. Pruning parameters, after the tree is built, the CART algorithm can also reduce the size of the tree by pruning to avoid overfitting.

[0078] Using the cleaned data set, the optimal features are selected for splitting based on the Gini index to construct a decision tree;

[0079] It should be noted that the Gini index is used to evaluate the degree to which samples in a subset belong to the same category under the split of a certain feature. The expression for calculating the Gini index is:

[0080]

[0081] Where G(D) represents the Gini index of the user behavior dataset D; D represents the user behavior dataset; f represents the total number of maintenance personnel categories; i represents the index of the maintenance personnel category; p i represents the probability of maintenance personnel category i;

[0082] It should also be noted that the value of the Gini index is between [0, 1], and the smaller the value, the higher the purity.

[0083] Use cross-validation to evaluate the decision tree model, adjust hyperparameters and perform pruning to optimize the decision tree model;

[0084] It should be noted that the purpose of adjusting hyperparameters is to optimize the model's accuracy, stability, and generalization ability by controlling the complexity of the tree. Maximum depth controls the maximum depth of the tree, preventing it from being too deep and reducing overfitting. Minimum sample splits: The minimum number of samples required to split a node. Increasing this parameter can prevent the model from becoming too complex. Minimum sample leaves: The minimum number of samples required to be included in a leaf node. Increasing this parameter can make the tree branch more relaxed, thereby reducing overfitting. Maximum number of features: The number of features selected when splitting a node. Reducing the number of features selected can help reduce overfitting.

[0085] Apply the trained decision tree model to user behavior data, output the prediction results, and generate a decision tree model;

[0086] It should be noted that a decision tree performs reasoning along the tree's nodes based on the characteristics of new data. The model starts at the root node of the tree and determines which branch of the tree the data belongs to based on the splitting conditions of the features, eventually reaching a leaf node.

[0087] Input the pre-processed user behavior data into the decision tree model, make predictions, and generate prediction work orders.

[0088] It should be noted that the output of the predicted work order may include a detailed description of the work order, its urgency, the equipment or system involved, the estimated repair time, and the required resources. This information is used to guide subsequent work order allocation and resource scheduling.

[0089] The allocation result generation module uses big data analysis to process the predicted work orders based on the predicted work orders, obtains the priority ranking of the work orders handled by maintenance personnel, and uses the particle swarm optimization algorithm to schedule resources, optimize the predicted work order allocation, and generate the work order allocation results.

[0090] Based on the predicted work order, collect the data of the predicted work order and the information of the maintenance personnel;

[0091] It should be noted that the data used to predict work orders includes work order type, urgency, historical maintenance records, equipment status, user preferences, and maintenance personnel information. Maintenance personnel information includes skill rating, available tools, work location data, and current work status.

[0092] Quantify and comprehensively sort the predicted work order data, and classify the predicted work orders into three levels of urgency;

[0093] It should be noted that based on the comprehensive ranking results, the system categorizes work orders into three priority levels: high, medium, and low. High-priority work orders typically involve critical equipment functional failures, urgent repair needs, or strong user complaints. Medium-priority work orders are generally non-urgent issues that require timely repair. Low-priority work orders typically include routine inspections, non-urgent minor fault repairs, or preventive maintenance work.

[0094] Based on the maintenance personnel's skills and the predicted fault type of the work order, a scoring algorithm is used to score and obtain the maintenance personnel's maintenance skill priority;

[0095] It should be noted that a series of scoring criteria are defined based on maintenance personnel's skills, experience, certifications, past maintenance records, and problem-solving efficiency. Based on the scoring criteria and weights, a skill score is calculated for each maintenance personnel. Maintenance personnel are sorted in descending order based on their skill scores to determine their skill priority. Personnel with higher skill scores are ranked first, prioritizing complex or difficult work orders.

[0096] Based on the maintenance personnel's maintenance tools and predicted work order maintenance needs, a scoring algorithm is used to score and obtain the priority of the maintenance personnel's available tools;

[0097] If necessary, collect information about the tools used by each maintenance technician. This includes tool type, frequency of use, maintenance history, and tool wear. Score each tool according to pre-set criteria. Calculate a comprehensive score for each maintenance technician's tools. Rank each maintenance technician's tools based on their comprehensive tool scores.

[0098] Based on the location information of the maintenance personnel and the predicted work order location, the Euclidean distance between the maintenance personnel and the predicted work order location is calculated and evaluated to obtain the distance priority of the maintenance personnel;

[0099] It should be noted that the current location coordinates of each maintenance worker are obtained, typically in latitude and longitude or planimetric coordinates. The specific location coordinates of each pending work order are obtained. The calculated Euclidean distance is reverse-converted into a priority score. A closer distance means the maintenance worker can reach the work order location faster, thus receiving a higher priority.

[0100] A comprehensive evaluation is performed on the maintenance personnel's maintenance skill priority, the maintenance personnel's available tool priority, and the maintenance personnel's distance priority to obtain the maintenance personnel's priority;

[0101] It should be noted that each maintenance worker's overall priority score is obtained by multiplying the score of each factor by its weight using a weighted average method. All maintenance workers are ranked based on their overall priority scores, and the maintenance worker with the highest score is considered the most appropriate person to handle the work order.

[0102] Comprehensively evaluate and rank the predicted urgency of work orders and the priority of maintenance personnel to obtain the priority of maintenance personnel in handling work orders;

[0103] It should be noted that high-urgency work orders are processed promptly without delays due to maintenance personnel capabilities or distance. Maintenance personnel are able to handle work orders according to their priority, ensuring they can complete their tasks efficiently.

[0104] The initial state of the particle swarm is set according to the maintenance personnel's work order processing priority, and the initial work order allocation plan is randomly generated;

[0105] It should be noted that the maintenance worker priority here is the result of a comprehensive evaluation of multiple factors, including skills, tools, distance, and availability, representing each maintenance worker's ability and efficiency in handling a specific work order. When the particle swarm is initialized, each particle's position corresponds to a work order assignment scheme. This scheme is randomly generated, meaning that a maintenance worker is randomly assigned to each work order based on the current work order priority and maintenance worker priority. This initial work order assignment scheme may not be the optimal solution, but it provides a starting point for the subsequent particle swarm optimization process.

[0106] Based on the initial work order allocation plan, the quality of the work order allocation plan is evaluated through the fitness function;

[0107] It should be noted that work orders should be assigned and processed as soon as possible based on their urgency and processing time. Therefore, the allocation plan should consider the urgency of the work order to ensure that urgent work orders are handled promptly. Maintenance personnel skills should be matched. Each maintenance personnel has different skills and tools. An effective allocation plan should base allocation on the matching of maintenance personnel's skills and tools with the work order requirements. The distance between the maintenance personnel and the work order location is also a key factor.

[0108] Update the particle solution based on the quality of the work order allocation solution and the feedback of the fitness function;

[0109] It should be noted that the influence of a particle's previous velocity on its current velocity helps the particle maintain a certain search direction and momentum. Particles move closer to their historical optimal position. Particles hope to find a better solution than their current one, so they move toward their personal optimal solution. This movement toward the group optimal solution helps particles move toward the global optimal solution, avoiding being trapped in a local optimum.

[0110] Based on the solution of the updated particles, the global optimal solution is selected as the final allocation solution through the fitness function;

[0111] It should be noted that in the specific application of work order allocation, particles represent different work order allocation schemes, and the fitness value of each particle reflects the pros and cons of the scheme. The particle with the best fitness is selected as the global optimal solution, indicating that the allocation scheme is the most appropriate.

[0112] Based on the priority of the maintenance personnel in handling the work order, it is assigned to the maintenance personnel and the work order assignment result is generated.

[0113] It should be noted that the most suitable maintenance personnel should be identified. The most suitable personnel should be selected to handle a specific work order based on their skills, tools, and location. Workload balancing should be considered to ensure that the workload of each maintenance personnel is relatively even, so that some maintenance personnel are not overworked while others are idle. Task prioritization should be considered to assign urgent tasks to the appropriate personnel based on the urgency of the work order or the maintenance personnel's priority.

[0114] Based on the solution of the updated particles, the work order allocation plan is input into the fitness function for evaluation to obtain the fitness value of the work order allocation plan;

[0115] It should be noted that a higher fitness value indicates a better work order allocation plan, meaning that the rationality of work order allocation, urgency response, time efficiency, maintenance personnel load, and skill matching have all reached optimal levels. Conversely, a low fitness value indicates that the allocation plan has flaws and may need to be adjusted and optimized.

[0116] Use roulette wheel selection to select the solution with the best fitness value from the current solution set;

[0117] It should be noted that roulette wheel selection determines the probability of a solution being selected based on its fitness value, thereby ensuring that solutions with high fitness are selected more frequently, pushing the optimization process towards more optimal solutions. Based on the random number, its position among all solutions is calculated, and the solution to be selected is determined. The specific method is to distribute the fitness values ​​of the solutions proportionally on the roulette wheel, forming a range. Then, the solution is selected by comparing the range in which the random number falls.

[0118] The solution with the best fitness value is selected as the parent, and crossover and mutation operations are performed to generate a new work order allocation solution.

[0119] It should be noted that the crossover point is randomly selected, and generally a certain position is selected to exchange the genes of the two parents, that is, part of the information on work order allocation. The genes of the parents are exchanged at the crossover point to generate a new offspring solution. The purpose of crossover is to combine the advantages of the parent solutions to create a better new solution. One or more work orders are randomly selected to decide which parts to mutate. Adjustments to the selected work orders can change information such as the work order allocation, priority, or assigned maintenance personnel. The new work order allocation scheme generated by the crossover and mutation operations will become part of the next generation solution and participate in the subsequent evaluation and selection process. As the genetic algorithm iterates, the system will continuously optimize the solution set according to the fitness function, and ultimately arrive at the optimal work order allocation scheme.

[0120] Input the new work order allocation plan into the fitness function and calculate its fitness value;

[0121] It should be noted that based on the evaluation results, the fitness function will generate a specific fitness value, usually a numerical value, that indicates the quality of the current work order allocation scheme. A high fitness value indicates a better scheme, while a low fitness value indicates a worse scheme.

[0122] Compare the new fitness value with the old fitness value and select the solution with higher fitness as the global optimal solution;

[0123] It should be noted that the fitness of the current work order allocation solution is first evaluated to obtain the fitness value of the current solution. Assume that the fitness value of the current solution is A. The fitness of the newly generated work order allocation solution is then evaluated to obtain the fitness value of the new solution. Assume that the fitness value of the new solution is B. If A is greater than B, the new solution is superior to the current solution, and the system selects the new solution as the current optimal solution. If A is less than B, the current solution remains unchanged and is still the optimal solution.

[0124] The work order allocation plan with the global optimal solution is used as the final allocation plan.

[0125] It should be noted that through multiple generations of optimization in the genetic algorithm, the solution finally selected has the best fitness, which means that it can maximize the use of resources and allocate them most efficiently in the process of processing work orders.

[0126] The maintenance report generation module allows maintenance personnel to dynamically adjust the work order execution progress and generate maintenance reports through mobile terminal devices based on the work order allocation results.

[0127] Based on the work order assignment results, maintenance personnel receive the task and use mobile devices to update the maintenance progress in real time;

[0128] It should be noted that when a maintenance worker receives a work order assigned by the system via a mobile device, the device displays detailed information about the task, including the fault description, repair location, and required tools and materials. Once the maintenance worker confirms acceptance of the task, the system updates the work order status to "In Progress."

[0129] Predict work order completion time and progress based on real-time uploaded data, use the Kalman filter algorithm to dynamically adjust the work order execution progress and address any delays;

[0130] It should be noted that a suitable dynamic model is needed to describe the state changes of each task during the work order execution process. The model typically includes information such as the task's initial time, the estimated completion time of each stage, and the current progress. A Kalman filter is used to calculate the completed tasks and elapsed time based on the real-time data uploaded by maintenance personnel. This updates the task state estimate and error estimate, thereby predicting the remaining work time and estimated completion time of the work order.

[0131] When the maintenance progress does not match the expected time, the Kalman filter algorithm will automatically adjust the expected completion time and perform resource scheduling;

[0132] It should be noted that the Kalman filter algorithm automatically adjusts the prediction of the work order completion time based on the deviation between the real-time uploaded maintenance progress data and the original estimated time. The Kalman filter combines the system's dynamic model, the task's historical execution status, resource consumption, and current observations of the current progress and working time to dynamically adjust the task's remaining time.

[0133] The maintenance personnel completes the work order task and confirms the maintenance status as completed, and a maintenance report is automatically generated.

[0134] The repair report should include a repair overview, briefly describing the repair work order, the cause of the problem, and the repair objectives. Repair details, detailing the steps taken, materials, and tools used during the repair process. Repair results, including whether the equipment returned to normal operation after the repair and any potential issues. Additional information includes the repair technician's signature, repair time, and user feedback.

[0135] Track the completion of tasks in real time based on the initial estimated time of the work order and the actual uploaded progress data;

[0136] It should be noted that during the maintenance process, maintenance personnel use mobile devices such as smartphones and tablets to record and upload the actual progress data of the task in real time. This data may include completed maintenance steps, actual time taken, and current status.

[0137] By regularly calculating the actual completion time of the current maintenance step and comparing it with the set estimated time, the probability of delay is assessed;

[0138] It should be noted that by analyzing past delay data for similar tasks, the probability of delay for the current task under similar conditions is estimated. By combining current progress with historical data, a Kalman filter is used to dynamically predict future progress, thereby calculating the likelihood and extent of delays. A regression model is developed to correlate task delays with potentially influencing factors such as maintenance personnel efficiency and equipment complexity, thereby assessing the potential for delays in future steps.

[0139] According to the delay, the variance of the measurement error and the process error is determined to obtain the Kalman gain;

[0140] It should be noted that the Kalman gain is a trade-off factor that determines the degree of fusion between the model prediction value and the actual measurement value. The expression is:

[0141]

[0142] Among them, H k represents the Kalman gain, the weight of the influence of the current k-time measurement on the prediction; P k-1 represents the error covariance at time k-1; R represents the error variance of the current measurement; k represents the current time.

[0143] The Kalman gain adjusts the degree of correction to the predicted value based on the relative magnitude of the measurement error and process error. If the measurement error is small, the Kalman gain will be large, and the system will rely more on the current measurement data. If the process error is large, the Kalman gain will be small, and the system will rely more on the previous prediction results.

[0144] Update the estimated completion time of the work order task based on the deviation between the Kalman gain and the actual completion time;

[0145] It should be noted that the Kalman gain determines the weight of the actual completion time deviation on the estimated completion time correction, and the expression is:

[0146]

[0147] in, represents the estimated completion time at time k; Indicates the predicted completion time at time k-1; actual completion time is the deviation, which represents the difference between the actual progress and the predicted progress.

[0148] Automatically find other idle maintenance personnel to complete unfinished work order tasks through the scheduling algorithm.

[0149] It should be noted that the scheduling algorithm can automatically detect when a work order is behind schedule and quickly find other suitable maintenance personnel to take over the unfinished task based on certain rules. The selection of available personnel requires a comprehensive consideration of multiple factors, such as maintenance personnel skills, distance, and tool availability, to ensure the successful completion of the task.

[0150] The scoring and feedback module pushes work order completion notifications to users based on the completion of repairs and obtains user scoring and feedback data.

[0151] Automatically generate a work order completion push report based on the maintenance report, and push the work order completion to the user according to the notification receiving method pre-selected by the user;

[0152] It should be noted that the work order completion push report includes the work order number and description, maintenance personnel information, maintenance task completion status, maintenance results and maintenance date.

[0153] And push a link to the user feedback form, where users can fill in service ratings, maintenance personnel performance, problem resolution status, and suggestions;

[0154] It should be noted that users are asked to rate the overall quality of the maintenance service, evaluating the timeliness of the service and the professionalism of the maintenance personnel. Users will rate the maintenance personnel's attitude, professionalism, and problem-solving ability. Users can also assess whether the maintenance completely resolved their problem and achieved their expected results. Users can also provide specific suggestions and feedback to help the property management company improve service quality in the future.

[0155] The collected user feedback data is stored in the database, and user ratings and feedback data are obtained through analysis.

[0156] It should be noted that based on feedback data regarding slow maintenance progress and poor service attitudes, work order processing, maintenance personnel scheduling, and resource allocation will be optimized. For maintenance personnel with low ratings, targeted training will be provided to improve their skills and service attitudes. If users widely report issues with certain equipment and technologies, the system can trigger notifications for equipment maintenance and technology upgrades.

[0157] In summary, the present invention achieves this by: constructing a decision tree model to predict pre-processed user behavior data and generating predicted work orders, thereby improving the accuracy of work order generation, reducing manual judgment errors, and foreseeing potential problems in advance, thus avoiding response delays and uneven processing. Subsequently, based on the predicted work orders, a particle swarm optimization algorithm is used to schedule and allocate resources for the work orders, comprehensively considering factors such as maintenance personnel skills, tools, and distance to ensure optimal work order allocation. This optimization algorithm improves resource scheduling efficiency, reduces work order delays, optimizes the workload of maintenance personnel, and significantly improves the efficiency and quality of work order processing, thereby improving the overall effectiveness of property management services.

[0158] It should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and are not intended to limit the present invention. Although the present invention has been described in detail with reference to the preferred embodiments, those skilled in the art should understand that the technical solutions of the present invention may be modified or replaced by equivalents without departing from the spirit and scope of the technical solutions of the present invention, which should all be included in the scope of the claims of the present invention.

Claims

1. A property management work order acceptance service system, characterized by: include, The collection and preprocessing module collects user behavior data and preprocesses the user behavior data; The prediction work order generation module builds a decision tree model based on pre-processed user behavior data to make predictions and generate prediction work orders; The allocation result generation module uses big data analysis to process the predicted work orders based on the predicted work orders, obtains the priority ranking of the work orders handled by maintenance personnel, and uses the particle swarm optimization algorithm to perform resource scheduling, optimize the predicted work order allocation, and generate the work order allocation results; The maintenance report generation module allows maintenance personnel to dynamically adjust the work order execution progress and generate maintenance reports through mobile terminal devices based on the work order allocation results; The scoring and feedback module pushes work order completion notifications to users based on the completion of repairs and obtains user scoring and feedback data.

2. The property management work order acceptance service system according to claim 1, characterized in that: The user behavior data includes device usage frequency, historical fault records, current, voltage, temperature and humidity, and user repair report records.

3. The property management work order acceptance service system according to claim 2, characterized in that: The user behavior data is pre-processed as described above. The specific steps are as follows: Use Gaussian filter to filter noise from collected user behavior data; Detect and delete duplicate records through deduplication algorithms; Use correlation analysis and mutual information methods to filter features related to fault occurrence and work order priority; The data scale is unified by standardizing the numerical features.

4. The property management work order acceptance service system according to claim 3, characterized in that: Based on the pre-processed user behavior data, a decision tree model is constructed to make predictions and generate prediction work orders. The specific steps are as follows: Select the CART algorithm according to the nature of the task and set the decision tree model parameters; Using the cleaned data set, the optimal features are selected for splitting based on the Gini index to construct a decision tree; Use cross-validation to evaluate the decision tree model, adjust hyperparameters and perform pruning to optimize the decision tree model; Apply the trained decision tree model to user behavior data, output the prediction results, and generate a decision tree model; Input the pre-processed user behavior data into the decision tree model, make predictions, and generate prediction work orders.

5. The property management work order acceptance service system according to claim 4, characterized in that: Based on the predicted work order, the predicted work order is processed using big data analysis to obtain the priority ranking of the maintenance personnel to handle the work order. The specific steps are as follows: Based on the predicted work order, collect the data of the predicted work order and the information of the maintenance personnel; Quantify and comprehensively sort the predicted work order data, and classify the predicted work orders into three levels of urgency; Based on the maintenance personnel's skills and the predicted fault type of the work order, a scoring algorithm is used to score and obtain the maintenance personnel's maintenance skill priority; Based on the maintenance personnel's maintenance tools and predicted work order maintenance needs, a scoring algorithm is used to score and obtain the priority of the maintenance personnel's available tools; Based on the location information of the maintenance personnel and the predicted work order location, the Euclidean distance between the maintenance personnel and the predicted work order location is calculated and evaluated to obtain the distance priority of the maintenance personnel; A comprehensive evaluation is performed on the maintenance personnel's maintenance skill priority, the maintenance personnel's available tool priority, and the maintenance personnel's distance priority to obtain the maintenance personnel's priority; The predicted urgency of the work order and the priority of the maintenance personnel are comprehensively evaluated and ranked to obtain the priority of the maintenance personnel in handling the work order.

6. The property management work order acceptance service system according to claim 5, characterized in that: The particle swarm optimization algorithm is used to perform resource scheduling, optimize and predict work order allocation, and generate work order allocation results. The specific steps are as follows: The initial state of the particle swarm is set according to the maintenance personnel's work order processing priority, and the initial work order allocation plan is randomly generated; Based on the initial work order allocation plan, the quality of the work order allocation plan is evaluated through the fitness function; Update the particle solution based on the quality of the work order allocation solution and the feedback of the fitness function; Based on the solution of the updated particles, the global optimal solution is selected as the final allocation solution through the fitness function; Based on the priority of the maintenance personnel in handling the work order, it is assigned to the maintenance personnel and the work order assignment result is generated.

7. The property management work order acceptance service system according to claim 6, characterized in that: Based on the solution of the updated particles, the global optimal solution is selected as the final allocation solution through the fitness function. The specific steps are as follows: Based on the solution of the updated particles, the work order allocation plan is input into the fitness function for evaluation to obtain the fitness value of the work order allocation plan; Use roulette wheel selection to select the solution with the best fitness value from the current solution set; The solution with the best fitness value is selected as the parent, and crossover and mutation operations are performed to generate a new work order allocation solution. Input the new work order allocation plan into the fitness function and calculate its fitness value; Compare the new fitness value with the old fitness value and select the solution with higher fitness as the global optimal solution; The work order allocation plan with the global optimal solution is used as the final allocation plan.

8. The property management work order acceptance service system according to claim 7, characterized in that: According to the work order allocation results, maintenance personnel dynamically adjust the work order execution progress and generate maintenance reports through mobile terminal devices. The specific steps are as follows: Based on the work order assignment results, maintenance personnel receive the task and use mobile devices to update the maintenance progress in real time; Predict work order completion time and progress based on real-time uploaded data, use the Kalman filter algorithm to dynamically adjust the work order execution progress and address any delays; When the maintenance progress does not match the expected time, the Kalman filter algorithm will automatically adjust the expected completion time and perform resource scheduling; The maintenance personnel completes the work order task and confirms the maintenance status as completed, and a maintenance report is automatically generated.

9. The property management work order acceptance service system according to claim 8, characterized in that: When the maintenance progress does not match the expected time, the Kalman filter algorithm will automatically adjust the expected completion time and perform resource scheduling. The specific steps are as follows: Track the completion of tasks in real time based on the initial estimated time of the work order and the actual uploaded progress data; By regularly calculating the actual completion time of the current maintenance step and comparing it with the set estimated time, the probability of delay is assessed; According to the delay, the variance of the measurement error and the process error is determined to obtain the Kalman gain; Update the estimated completion time of the work order task based on the deviation between the Kalman gain and the actual completion time; Automatically find other idle maintenance personnel to complete unfinished work order tasks through the scheduling algorithm.

10. The property management work order acceptance service system according to claim 9, characterized in that: Based on the maintenance report, the work order completion notification is pushed to the user, and the user rating and feedback data are obtained. The specific steps are as follows: Automatically generate a work order completion push report based on the maintenance report, and push the work order completion to the user according to the notification receiving method pre-selected by the user; And push a link to the user feedback form, where users can fill in service ratings, maintenance personnel performance, problem resolution status, and suggestions; The collected user feedback data is stored in the database, and user ratings and feedback data are obtained through analysis.

Images