Property intelligent resource scheduling optimization method based on Internet of Things

Abstract

The invention belongs to the technical field of resource scheduling, and provides a property intelligent resource scheduling optimization method based on the Internet of Things, and the method comprises the steps: obtaining contradictory event tags, events and association relationships of active feedback and passive perception of an owner and a property, and constructing a contradictory event data set; constructing a contradictory path model, inputting a data set to extract contradictory conduction paths, and performing similar combination to obtain multivariate and original contradictory paths; extracting property resource allocation imbalance characteristics when the multivariate contradictory paths are combined, and identifying whether a superimposed effect exists or not; if yes, screening the multi-element contradictory path area, and extracting a high-frequency superposition area; and inputting an intelligent resource allocation model on the basis of the high-frequency superposition region and the real-time passive perception characteristics, thereby realizing intelligent prediction and allocation optimization of the unbalanced property resources. According to the invention, by integrating multi-source data and intelligent analysis, the property resource scheduling efficiency is improved, the probability of occurrence of contradictory events is reduced, and the method has the advantages of prospective and intelligent resource allocation.

Description

Technical Field

[0001] The present invention belongs to the technical field of resource scheduling, and in particular to a property intelligent resource scheduling optimization method based on the Internet of Things. Background Art

[0002] With the popularization of IoT technology and the advancement of smart community construction, property management is increasingly demanding intelligent and refined resource scheduling. Traditional property resource scheduling relies primarily on manual feedback and experience-driven decision-making, which has many shortcomings.

[0003] Traditional scheduling techniques lack a deep connection between conflicting events and resource allocation, failing to analyze conflict transmission pathways. For example, how aging equipment can lead to delayed maintenance responses, which in turn triggers owner dissatisfaction. Resource allocation is often based on static rules, ignoring the transmission effects of different conflicting events and the cumulative impact of resource imbalances, resulting in inefficient scheduling.

[0004] Existing technologies lack the ability to integrate and analyze multi-source data, making it difficult to quantify resource allocation imbalances in real time and predict their trends. Property resource scheduling relies on manual experience, failing to proactively optimize areas of high-frequency conflicts. This often leads to a vicious cycle of passively responding to complaints and then temporarily adjusting resources, limiting improvements in owner satisfaction.

[0005] To this end, the present invention provides a property intelligent resource scheduling optimization method based on the Internet of Things. Summary of the Invention

[0006] In order to make up for the deficiencies of the prior art, at least one technical problem raised in the background technology is solved.

[0007] The technical solution adopted by the present invention to solve the technical problem is: a property intelligent resource scheduling optimization method based on the Internet of Things, comprising:

[0008] A property intelligent resource scheduling optimization method based on the Internet of Things includes the following steps:

[0009] Obtain conflict event labels and conflict events actively fed back and passively perceived by owners and property management, as well as conflict correlation relationships, to build a conflict event dataset;

[0010] Construct a contradiction path model, input the contradiction event data set into the contradiction path model, extract the contradiction transmission path and conduct similarity merging analysis to obtain multiple contradiction paths;

[0011] Extract the imbalance characteristics of property resource allocation when multiple conflict paths are merged, and identify whether the imbalance characteristics of property resource allocation have a superposition effect;

[0012] If there is a superposition effect, conduct regional screening analysis on the multiple conflict paths and extract the high-frequency superposition areas of all the multiple conflict paths;

[0013] Based on the determined high-frequency superposition area, the passive perception features collected in real time are input into the intelligent resource allocation model to realize intelligent prediction of unbalanced property resources and optimize the allocation of property resources.

[0014] Furthermore, the method of constructing the conflict event dataset is as follows:

[0015] Obtain the conflict event label and the event resources and property maintenance resources in the conflict event process, and perform correlation matching to obtain the conflict correlation relationship of the conflict event;

[0016] The conflict event label, event resources in the conflict event process, and property maintenance resources are used as conflict event elements;

[0017] Obtain the conflict event elements and conflict correlation relationships of all conflict events between property management and owners to build a conflict event dataset.

[0018] Furthermore, the contradictory path model is as follows:

[0019] Perform data preprocessing on the input conflict event data set;

[0020] Construct a contradiction association graph based on the contradiction event dataset after data preprocessing;

[0021] The contradiction transmission path is extracted based on the contradiction association graph.

[0022] Furthermore, the similarity merging analysis method is:

[0023] Extract any two contradictory conduction paths as a path comparison group, extract the two longest common subpaths of the path comparison group, and calculate the length of any two longest common subpaths;

[0024] Calculate the similarity of the lengths of any two longest common subpaths. Obtain the similarity of any two longest common subpaths in the path comparison group;

[0025] If the similarity of the longest common sub-path is higher than a preset similarity threshold, the contradictory conduction paths of the path comparison group are merged through a clustering algorithm.

[0026] Furthermore, the method for identifying whether the imbalanced allocation of property resources has a superposition effect is:

[0027] Based on the imbalance characteristics of property resource allocation when multiple conflict paths are merged, a vector autoregression model is constructed to conduct autocorrelation analysis on the imbalance characteristics of property resource allocation of multiple conflict paths.

[0028] If the impact coefficient of the autocorrelation analysis is significantly different from zero, the imbalance characteristics of property resource allocation have a superimposed effect.

[0029] Furthermore, the method for obtaining the property resource allocation imbalance feature is:

[0030] From the merging process of multiple contradictory paths, the original contradictory paths initially corresponding to the multiple contradictory paths are obtained;

[0031] Identify the property resource imbalance characteristics that are filtered out when multiple conflicting paths are merged, and add the filtered imbalance characteristics to the path mapping table;

[0032] The spatial mismatch rate and time lag rate are extracted from the path mapping table as the imbalance characteristics of property resource allocation.

[0033] Furthermore, the spatial mismatch rate is obtained as follows:

[0034] The spatial mismatch rate is obtained by numerically calculating the density of maintenance personnel in the conflict-occurring area in the original conflict path and the maintenance personnel density and the actual maintenance personnel coverage radius.

[0035] Furthermore, the time lag rate is obtained as follows:

[0036] The time lag rate is obtained by numerically calculating the standardized time deviation value of the events in the original contradiction path.

[0037] Furthermore, the method of extracting the high-frequency superposition area of all multiple conflict paths is:

[0038] For multiple conflict paths with superposition effects, the conflict occurrence areas of all multiple conflict paths with superposition effects are summarized;

[0039] The frequencies of occurrence of various conflict regions of the multi-contradiction paths in different cycles are calculated, and high-frequency superposition regions are determined based on the frequencies of occurrence of the conflict regions in different cycles.

[0040] Furthermore, the method of optimizing the allocation of property resources is:

[0041] Based on the determined high-frequency superposition area and the passive sensing features collected in real time by IoT devices;

[0042] The high-frequency superposition areas and passive perception features collected in real time by IoT devices are input into the intelligent resource allocation model to achieve intelligent prediction of unbalanced property resources and optimize the allocation of property resources.

[0043] The beneficial effects of the present invention are as follows:

[0044] 1. Capture multi-dimensional conflict events and collect conflicts across all channels to form the data foundation for intelligent scheduling: By combining active feedback with passive perception, explicit complaints and implicit conflicts are converted into structured conflict event labels, reducing data blind spots in a single channel; based on IoT positioning technology, conflict events are dynamically bound to the occurrence area, equipment status, and maintenance resource elements to form multi-dimensional data associations, providing full-process traceable data support for subsequent path analysis.

[0045] 2. Analyze the conflict transmission paths to achieve differentiated processing of high-frequency problems and individual problems. By constructing a conflict association graph and analyzing the longest common sub-path, similar conflict transmission paths are merged to generate "multiple conflict paths" representing high-frequency common patterns, filtering out low-frequency atypical events, and prioritizing resource scheduling for repetitive and large-scale problems to improve optimization efficiency. It distinguishes between "original conflict paths" and "multiple conflict paths", reduces the incidence of common problems through large-scale processing, provides an identification basis for individual resource allocation problems, and takes into account both efficiency and flexibility of problem optimization.

[0046] 3. Quantitative analysis of superposition effects to locate key areas for resource allocation. Spatial mismatch rates and time lag rates are extracted from the process of merging conflicting paths. The autocorrelation and cross-influence of spatial mismatch rates and time lag rates are quantified through a vector autoregression model to identify the superposition effects of single or multi-dimensional imbalances and reduce scheduling decision-making biases caused by isolated analysis. Based on the results of the superposition effect analysis, periodic frequency statistics are used to screen superposition areas where conflicting events frequently occur. Combining historical data with real-time perception data, resource scheduling is focused on core areas that truly require optimization, reducing the inefficiency and waste caused by average resource allocation.

[0047] 4. Forward-looking intelligent prediction and dynamic allocation, realizing the transition from passive response to active optimization, inputting high-frequency superimposed regional characteristics and real-time IoT data into the intelligent resource allocation model, learning historical cycle patterns and real-time abnormal fluctuations, predicting imbalances such as maintenance personnel shortages and spare parts shortages in advance, and outputting forward-looking resource optimization strategies to resolve conflicts in the bud; by dynamically optimizing maintenance personnel task allocation and spare parts inventory layout, the spatiotemporal distribution of property resources is highly matched with actual demand, thereby improving the efficiency of property intelligent resource scheduling. BRIEF DESCRIPTION OF THE DRAWINGS

[0048] The present invention will be further described below with reference to the accompanying drawings.

[0049] Figure 1 This is a flow chart of a property intelligent resource scheduling optimization method based on the Internet of Things according to an embodiment of the present invention;

[0050] Figure 2 is a flow chart of the contradictory path model according to an embodiment of the present invention;

[0051] Figure 3 This is a module diagram of a property intelligent resource scheduling and optimization system based on the Internet of Things in an embodiment of the present invention. DETAILED DESCRIPTION

[0052] In order to make the technical means, creative features, objectives and effects achieved by the present invention easier to understand, the present invention is further described below in conjunction with specific implementation methods.

[0053] Example 1

[0054] See also Figure 1 As shown, the method for optimizing property intelligent resource scheduling based on the Internet of Things according to an embodiment of the present invention includes the following steps:

[0055] Step 1: Obtain conflict event labels and conflict events actively fed back and passively perceived by owners and property management, as well as conflict correlation relationships, and construct a conflict event dataset;

[0056] In some embodiments, a feedback communication channel between the owner and the property is established, the feedback communication channel is used to obtain the owner's feedback, and conflicting events between the owner and the property are identified in the feedback;

[0057] It needs to be explained that the identification of contradictory events is mainly achieved through two channels: active feedback and passive perception;

[0058] Active feedback is provided through feedback communication channels such as the property management mobile client, customer service phone, and suggestion boxes. By combining technologies such as natural language processing (NLP) and optical image recognition (OCR), owners' explicit complaints are converted into conflict event labels. For example, negative reviews, phone recordings, and text suggestions are converted into conflict event labels such as "delayed maintenance response" and "poor service quality of workers."

[0059] Passive sensing collects IoT device data such as maintenance time and equipment failure frequency, and conducts correlation analysis with the maintenance work order system to capture potential conflicting events such as duplicate repair reports and resource misallocation.

[0060] Based on active feedback and passive perception of conflict event labels and conflict events, the conflict-generating area is located through IoT devices, and event resources consisting of physical objects, owners, and time in the conflict event process are obtained. In addition, the equipment maintenance rate, maintenance personnel, and maintenance time in the conflict-generating area are obtained as property maintenance resources;

[0061] The physical objects may be malfunctioning equipment, spare parts that need to be replaced, etc.

[0062] Based on the conflict event labels and the event resources and property maintenance resources in the conflict event process, the conflict association relationship of the conflict event is obtained;

[0063] The conflict event label, event resources in the conflict event process, and property maintenance resources are used as conflict event elements;

[0064] Obtain the conflict event elements and conflict correlation relationships of all conflict events between property management and owners to build a conflict event dataset;

[0065] Preferably, APP complaints, active feedback from owners via phone recordings, and passive sensing data recorded by equipment sensor data and work order systems are used to collect conflict event information. Keywords are extracted from conflict event information through NLP to generate conflict labels such as "service quality" and "response lag". IoT positioning is used to determine the area where the conflict occurred and associate event resources in the area. For example, on January 1, 2025, the complaint record of an elderly person living alone, numbered U-007, was filed on the 1st floor of Unit 2, Building 3, with elevator number T-0302 and elevator spare parts usage records.

[0066] The basic information of the conflict event, event resources, conflict labels and related evidence (maintenance worker skill mismatch, equipment aging data) are structured and integrated to form a conflict event element including "event - owner - physical object - maintenance - label";

[0067] Based on the conflict event dataset, it provides standardized data support for subsequent conflict analysis and resource scheduling optimization.

[0068] Step 2: Construct a contradiction path model. Input the contradiction event data set into the contradiction path model, extract the contradiction transmission paths, and perform similarity merging analysis to obtain multiple contradiction paths and the original contradiction path.

[0069] like Figure 2 As shown in the figure, the contradictory path model is constructed as follows:

[0070] A1. Preprocess the input conflict event dataset;

[0071] Preferably, data preprocessing includes three steps: data cleaning, data standardization, and data encoding;

[0072] Data cleaning: used to remove duplicate records, erroneous data, and missing values from contradictory event datasets;

[0073] For example, if the device number in a feedback is empty, that is, it is a missing value and cannot be supplemented by other information, then the feedback with the empty device number can be deleted;

[0074] Data standardization: Standardize different types of data in the conflict event dataset to facilitate subsequent analysis. For example, normalize numerical data such as equipment maintenance frequency and maintenance duration.

[0075] Data encoding: Convert conflict event labels and owner types into numerical data to facilitate subsequent calculations;

[0076] A2. Construct a conflict association graph based on the conflict event dataset after data preprocessing.

[0077] The association graph is constructed as follows:

[0078] Extract the contradictory event elements from the contradictory event data set after data preprocessing as path nodes, and use the contradictory association relationships as path edges of the contradictory event elements;

[0079] Establish an association graph based on path nodes and contradictory edges through the model association algorithm;

[0080] Preferably, the model algorithm can be a graph database Neo4j or NetworkX;

[0081] It should be explained that the path nodes can be owners, equipment, conflict event labels, maintenance personnel, spare parts, etc.

[0082] Path edges are conflict association relationships, which can be complaint relationships between owners and conflict events, failure relationships between equipment and conflict events, or handling relationships between maintenance personnel and owners.

[0083] The model association algorithm connects the path nodes and path edges of the conflict event dataset that has been preprocessed to establish a conflict association graph;

[0084] A3. Extract the contradiction transmission path based on the contradiction association graph;

[0085] Preferably, through a graph traversal algorithm, the owner is taken as the starting point of the conflict transmission path, and the conflict transmission path is extracted along the path edge;

[0086] It needs to be explained that the role of extracting the contradiction transmission path is:

[0087] Function 1: Identify high-frequency common conflict patterns. Through graph traversal algorithms and similarity merging analysis, extract high-frequency and recurring conflict transmission patterns (such as "equipment aging → delayed maintenance response → repeated complaints from owners") from a large number of conflict events, forming "multiple conflict paths". These paths represent large-scale and typical resource allocation problems, and are the core contradictions that need to be solved as a priority in property resource scheduling.

[0088] Function 2: Distinguish between common and individual contradictions. During the merging process, similar paths are merged into high-frequency common "multiple contradiction paths", while retaining unmerged low-frequency, unique paths as "original contradiction paths" (such as skill mismatch of maintenance personnel in a specific area, single sudden spare parts shortage). This allows resource scheduling to not only optimize high-frequency problems in batches, but also take into account personalized and occasional resource imbalance scenarios, thereby improving the efficiency of scheduling strategies.

[0089] Function 3: Providing a structured basis for scheduling decisions. The conflict transmission path clarifies the causal chain of conflicting events (such as the relationship between owner complaints and equipment failures and maintenance resource allocation), transforming unstructured conflict events into structured data containing "node-edge-attribute" (such as owner node → equipment failure edge → maintenance worker node). This structured information provides direct input for subsequent analysis of resource allocation imbalance characteristics (such as spatial mismatch rate and time lag rate) and the construction of intelligent scheduling models, making resource optimization strategies more targeted.

[0090] Conduct similarity and overlap analysis on the conflict transmission paths and merge similar conflict transmission paths;

[0091] Among them, the method of similarity coincidence analysis of contradiction transmission paths is as follows:

[0092] Extract any two contradictory conduction paths as a path comparison group, extract the two longest common subpaths of the path comparison group, calculate the lengths of any two longest common subpaths, and obtain len1 and len2;

[0093] Wherein, len1 represents the length of the longest common subpath of the first contradictory conduction path in the path comparison group, and len2 represents the length of the longest common subpath of the second contradictory conduction path in the path comparison group;

[0094] The similarity of any two longest common subpaths in the path comparison group is obtained through the formula: Sim = len1 / len2;

[0095] Preferably, the length of the longest common subpath is obtained by constructing a two-dimensional table dp using a dynamic programming method. AB [a][b],dp AB [a][b] represents the longest common subpath length between the first a nodes of the contradiction conduction path A and the first b nodes of the contradiction conduction path B:

[0096] The similarity of the two longest common subpaths is compared with a preset similarity threshold. If the similarity of the longest common subpath is higher than the preset similarity threshold, the conflicting conduction paths of the path comparison group are merged through a clustering algorithm.

[0097] It should be explained that when merging and processing similar conflict transmission paths through clustering algorithms, the path similarity is first calculated based on the length of the longest common subpath (LCS) (e.g., LCS length ratio ≥ 60%), and a similarity matrix is constructed. Then, hierarchical clustering or DBSCAN algorithms are used to divide the paths into several clusters according to a threshold. For each path within a cluster, common nodes (such as the same device, high-frequency conflict labels) and edges (such as complaint relationships, processing records) are retained. Attribute data (such as the cumulative number of owner complaints and the weighted average of the maintenance worker skill mismatch rate) are merged, and conflict information (such as different maintenance workers handling the same device) is marked as multi-node associations. Finally, a representative path containing core conflict elements is generated, which is used to refine high-frequency transmission patterns (such as "equipment aging → delayed maintenance response → repeated owner complaints") and provide structured input for resource scheduling optimization.

[0098] The conflict conduction paths processed by merging are marked as multiple conflict paths, and the conflict conduction paths not processed by merging are marked as original conflict paths;

[0099] Those skilled in the art will understand that the original conflict path represents a unique conflict transmission pattern that has not been clustered, and may include personalized resource allocation issues, such as equipment aging in a specific area and the mismatch between maintenance personnel skills, a single sudden resource shortage, etc.

[0100] Multiple conflict paths refer to common conflict patterns with high frequency, which have been combined to filter out low-frequency and atypical resource problems;

[0101] The technical solution of this embodiment is as follows: obtaining conflict event labels and conflict events actively fed back and passively perceived by owners and property management, as well as conflict correlation relationships, to construct a conflict event dataset; constructing a conflict path model, inputting the conflict event dataset into the conflict path model, extracting the conflict transmission paths and merging and analyzing them similarly to obtain multiple conflict paths and original conflict paths; filtering low-frequency atypical events so that resource scheduling prioritizes repetitive and large-scale issues, thereby improving optimization efficiency.

[0102] Example 2

[0103] like Figure 1 As shown, a property intelligent resource scheduling optimization method based on the Internet of Things also includes the following steps:

[0104] Step 3: Extract the imbalance characteristics of property resource allocation when multiple conflict paths are merged, and identify whether the imbalance characteristics of property resource allocation have an additive effect;

[0105] Among them, the method of identifying the imbalance characteristics of property resource allocation when multiple conflicting paths are merged is:

[0106] From the merging process of multiple contradictory paths, the original contradictory paths initially corresponding to the multiple contradictory paths are obtained;

[0107] It should be explained that the original conflict path is a unique conflict transmission mode that has not been clustered, such as personalized resource allocation problems, equipment aging in a specific area and maintenance personnel skill mismatch, single sudden resource shortage, etc., which represents a low-frequency, personalized conflict transmission path;

[0108] The multi-path conflict is a "representative path" generated by merging high-frequency, common original paths through similarity analysis. It filters out low-frequency, atypical resource problems and focuses on high-frequency conflict patterns, such as "aging equipment leading to delayed maintenance response and repeated complaints from owners."

[0109] Those skilled in the art will understand that the initial source of the multiple contradictory paths is the original contradictory paths in the historical data. The two are in a relationship from "individual to group" and "specific to general". The merging process conforms to the clustering logic of "from concrete to abstract" in data mining.

[0110] When the original conflicting paths are merged into multiple conflicting paths, the numbers and merging times of the original conflicting paths are recorded during cluster merging analysis, and a path mapping table from multiple conflicting paths to original path clusters is established;

[0111] For example, if the multi-contradiction path "equipment aging → delayed maintenance response → owner complaints" corresponds to 100 original conflict paths, 90 of which come from the equipment aging area and 10 involve maintenance personnel skill mismatch;

[0112] Identify the property resource imbalance characteristics that are filtered out when multiple conflicting paths are merged, and add the filtered imbalance characteristics to the path mapping table;

[0113] It should be explained that the way to add the filtered imbalance characteristics to the path mapping table is as follows: the multi-contradictory path filters out low-frequency and atypical original paths, which may imply imbalance characteristics such as local resource mismatch and sudden shortage:

[0114] Resource mismatch: This category shows the frequency of the event label "maintenance worker skills do not match equipment fault type" in the original conflict path. For example, a maintenance worker failed to resolve an elevator circuit fault three times, but the multi-contradiction path only retains the label "maintenance response delay," which does not reflect the skill mismatch.

[0115] Sudden shortages: The event label "single spare parts inventory shortage leading to maintenance delay" in the original conflict path is not merged because it occurs less frequently, but it may occur repeatedly in a specific area.

[0116] Extract spatial mismatch rate and time lag rate from the path mapping table as the imbalance characteristics of property resource allocation;

[0117] S1. The spatial mismatch rate is obtained as follows:

[0118] By formula: Obtain the spatial mismatch rate SMR;

[0119] Among them, ρ i is the maintenance worker density in the conflict-occurring region i in the original conflict path, that is, the ratio of the number of maintenance workers in the conflict-occurring region to the area of the region;

[0120] ρ avg The average value of the maintenance worker density in the multi-contradiction path is obtained by averaging the maintenance worker density of all relevant areas;

[0121] R i is the actual maintenance personnel coverage radius of the conflict-occurring region i in the original conflict path, R std The standard coverage radius is the theoretical coverage radius calculated based on the standard service area of a single maintenance worker set by the property.

[0122] i is the number of the area where the conflict occurs;

[0123] S2. The time lag rate is obtained as follows:

[0124] By formula: Get the time lag rate TLR;

[0125] Where M is the total number of event path work orders constructed based on active feedback and passive perception of conflicting events in the multi-contradiction path, and j is the number of the event path work order;

[0126] It should be explained that the event path work order is established based on the conflicting events and is used to assign tasks to maintenance personnel to resolve the conflicting events;

[0127] Z j It represents the standardized time deviation of the event in the original conflict path, which is calculated by the deviation ratio between the maintenance worker's maintenance time and the preset standard time T;

[0128] ΔT j The deviation value is calculated by calculating the difference between the maintenance time of the maintenance worker and the preset standard time T;

[0129] I is the indicator function, which is 1 when I meets the condition, otherwise it is 0;

[0130] Based on the imbalance characteristics of property resource allocation when multiple conflict paths are merged, determine whether the imbalance characteristics of property resource allocation of multiple conflict paths have a superposition effect;

[0131] Among them, the method to judge whether the imbalanced allocation characteristics of property resources in multiple conflict paths have a superposition effect is as follows:

[0132] By formula: A vector autoregression model was constructed to conduct autocorrelation analysis on the imbalance characteristics of property resource allocation in multiple conflict paths.

[0133] Among them, TLR t It represents the time lag rate at time t, reflecting the degree of deviation between the current maintenance time and the standard time;

[0134] α k TLR hysteresis effect k period t time TLR t The influence coefficient of TLR is used to measure the effect of its own historical value on the current value.

[0135] TLR t-k represents the value of TLR at time tk, that is, the historical observation value of TLR;

[0136] β k represents the effect of SMR after lag k on TLR at time t t The influence coefficient reflects the cross-effect of spatial mismatch rate on time lag rate;

[0137] SMR t-k represents the value of SMR at time tk, that is, the historical observation value of SMR;

[0138] ε 1t It represents the random error term of the TLR equation at time t, which obeys a normal distribution with a mean of 0 and captures the random factor TLR not explained by the model. t The influence of ε 2t It represents the random error term of the SMR equation at time t, which obeys a normal distribution with a mean of 0 and captures the random factor SMR not explained by the model. t the impact of;

[0139] SMR t It represents the spatial mismatch rate at time t, and quantifies the imbalance between the maintenance personnel density and coverage radius in the current area;

[0140] γ k :TLR lag k period to SMR at time t t The influence coefficient of reflects the reverse effect of time lag rate on spatial mismatch rate;

[0141] δ k :SMR lagged k period to SMR at time t t The influence coefficient of SMR describes the impact of its own historical value on the current value;

[0142] q is the lag order of the model, which determines how many past periods of TLR and SMR values are used to explain the current values. It needs to be determined based on the data characteristics and the AIC and BIC criteria of the statistical test;

[0143] t is the time, k is the number of lag periods;

[0144] What needs to be explained is that is the autoregressive part of the vector autoregressive model, which respectively represents the impact of the historical values of the time lag rate and spatial mismatch rate on the current value;

[0145] If α k or δ k If the value is significantly different from zero, it indicates that the imbalance feature has temporal autocorrelation, that is, the current imbalance level is affected by the past imbalance state, which reflects the superposition effect of a single imbalance feature in the time dimension.

[0146] For example, when α1 = 0.8, it means that the time lag rate of the previous period has a greater positive impact on the current time lag rate, and the imbalance of the previous period will continue to the current period, showing a superposition trend;

[0147] is the cross part in the formula, which represents the mutual influence between different imbalance characteristics; if β k or γ k If it is significantly different from zero, it indicates that different types of imbalance characteristics are correlated with each other in time;

[0148] For example, β1 = 0.6, which means that for every unit increase in the spatial mismatch rate in the previous period, the time lag rate in the current period will increase by 0.6 units. This reflects the superposition effect of different imbalance characteristics in the time dimension, that is, the deterioration of one imbalance characteristic will further aggravate the degree of another imbalance characteristic.

[0149] It should be explained that the role of identifying whether there is an additive effect is:

[0150] Function 1: Revealing the dynamic correlation of imbalance characteristics is conducive to reducing the possibility of isolated analysis. Traditional resource scheduling often analyzes a single imbalance problem in isolation (such as focusing only on the insufficient number of maintenance personnel or excessive maintenance time). The superposition effect identification quantifies the autocorrelation (the impact of historical imbalance on the current state) and cross-influence (the interaction between different characteristics) of different imbalance characteristics (such as spatial mismatch rate and time lag rate TLR) through the vector autoregression model (VAR).

[0151] Function 2: Identify high-frequency, complex imbalance areas and focus on core contradictions. The existence of a superposition effect usually means that multiple resource imbalance problems in a certain area are worsening in a coordinated manner (such as insufficient maintenance personnel, shortage of spare parts, and concentrated aging equipment). Such areas are often high-incidence areas of conflict incidents. By identifying the superposition effect and screening high-frequency superposition areas, resource allocation priorities can be achieved, and human and material resources can be invested first in high-frequency areas with a superposition effect.

[0152] Function three: Reduce the incidence of conflicting incidents, optimize resource utilization efficiency, reduce duplicate repairs and resource waste by identifying the superposition effect and optimizing it in a targeted manner, and improve resource utilization efficiency.

[0153] Step 4: If there is a superposition effect, perform regional screening analysis on the multiple conflict paths and extract the high-frequency superposition areas of all the multiple conflict paths;

[0154] If there is a superposition effect, screen the multiple contradiction paths with superposition effect among all the multiple contradiction paths;

[0155] The method of extracting the high-frequency superposition area of all multiple conflict paths is:

[0156] For multiple conflict paths with superposition effects, the conflict occurrence areas of all multiple conflict paths with superposition effects are summarized;

[0157] Calculate the frequency of each conflict occurrence area of the multi-contradiction path in different cycles, compare it with the preset frequency threshold, and mark the conflict occurrence area with a frequency higher than or equal to the preset threshold as a high-frequency superposition area;

[0158] Those skilled in the art will appreciate that the preset frequency threshold is obtained by those skilled in the art through statistical analysis of historical data.

[0159] Step 5: Based on the determined high-frequency superposition area, the passive sensing features collected in real time are input into the intelligent resource allocation model to achieve intelligent prediction of unbalanced property resources and optimize the allocation of property resources;

[0160] Based on the determined high-frequency superposition area and the passive sensing features collected in real time by IoT devices;

[0161] Inputting high-frequency overlapping areas and passive sensing features collected in real time by IoT devices into the intelligent resource allocation model enables intelligent prediction of unbalanced property resources, optimizes the allocation of property resources, and helps reduce the probability of conflict events.

[0162] It is understood by those skilled in the art that the construction of the intelligent resource allocation model requires high-frequency superimposed regional features and real-time passive perception data of the Internet of Things as core inputs. By integrating the imbalance characteristics of the spatiotemporal dimension (such as spatial mismatch rate (SMR) and time lag rate (TLR), equipment operating status (failure frequency, spare parts inventory) and maintenance resource distribution (maintenance personnel skills and location), a hybrid architecture of long short-term memory network (LSTM) combined with graph neural network (GNN) is adopted to capture historical cycle patterns and real-time abnormal fluctuations.

[0163] The intelligent resource allocation model learns the imbalance patterns in high-frequency areas through training (such as "nighttime + equipment aging area → insufficient maintenance personnel + spare parts shortage"), and outputs dynamic allocation strategies (personnel dispatch, spare parts pre-storage, preventive maintenance), ultimately achieving forward-looking optimization of resources such as maintenance personnel and spare parts, reducing the probability of conflict incidents.

[0164] The technical solution of this embodiment is as follows: extracting the imbalance characteristics of property resource allocation when multiple conflicting paths are merged, and identifying whether the imbalance characteristics of property resource allocation have an overlay effect; if there is an overlay effect, performing regional screening and analysis on the multiple conflicting paths, and extracting the high-frequency overlay areas of all the multiple conflicting paths; based on the determined high-frequency overlay areas, inputting the passive perception characteristics collected in real time into the intelligent resource allocation model to achieve intelligent prediction of unbalanced property resources and optimize the allocation of property resources; performing forward-looking intelligent prediction and dynamic allocation of unbalanced property resources, achieving a transition from passive response to active optimization, inputting the high-frequency overlay area characteristics and real-time data from the Internet of Things into the intelligent resource allocation model, learning historical cycle patterns and real-time abnormal fluctuations, and predicting imbalance states such as maintenance personnel shortages and spare parts shortages in advance, outputting forward-looking resource optimization strategies, and resolving conflicting events in the bud; by dynamically optimizing maintenance personnel task allocation and spare parts inventory layout, the spatiotemporal distribution of property resources is highly matched with actual demand, thereby improving the efficiency of intelligent property resource scheduling.

[0165] Example 3

[0166] like Figure 3 As shown in the figure, a property intelligent resource scheduling and optimization system based on the Internet of Things includes the following modules:

[0167] Conflict extraction module: used to obtain conflict event labels and events actively fed back and passively perceived by owners and property management, as well as conflict associations, and construct a conflict event dataset;

[0168] By establishing a feedback communication channel between owners and property management, we can obtain feedback from owners and identify conflicts between owners and property management within the feedback;

[0169] Based on active feedback and passive perception of conflict event labels and conflict events, the conflict-generating area is located through IoT devices, and event resources consisting of physical objects, owners, and time in the conflict event process are obtained. In addition, the equipment maintenance rate, maintenance personnel, and maintenance time in the conflict-generating area are obtained as property maintenance resources;

[0170] The physical objects may be malfunctioning equipment, spare parts that need to be replaced, etc.

[0171] Based on the conflict event labels and the event resources and property maintenance resources in the conflict event process, the conflict association relationship of the conflict event is obtained;

[0172] The conflict event label, event resources in the conflict event process, and property maintenance resources are used as conflict event elements;

[0173] Obtain the conflict event elements and conflict correlation relationships of all conflict events between property and owners to build a conflict event dataset.

[0174] Path merging module: used to build a contradiction path model, input the contradiction event data set into the contradiction path model, extract the contradiction transmission path and perform similarity merging analysis to obtain multiple contradiction paths and the original contradiction path;

[0175] Among them, the construction method of the contradictory path model is:

[0176] A1. Preprocess the input conflict event dataset;

[0177] Data cleaning: used to remove duplicate records, erroneous data, and missing values from contradictory event datasets;

[0178] Data standardization: Standardize different types of data in the conflict event dataset to facilitate subsequent analysis. For example, normalize numerical data such as equipment maintenance frequency and maintenance duration.

[0179] Data encoding: Convert conflict event labels and owner types into numerical data to facilitate subsequent calculations;

[0180] A2. Construct a conflict association graph based on the conflict event dataset after data preprocessing.

[0181] The association graph is constructed as follows:

[0182] Extract the contradictory event elements from the contradictory event data set after data preprocessing as path nodes, and use the contradictory association relationships as path edges of the contradictory event elements;

[0183] Establish an association graph based on path nodes and contradictory edges through the model association algorithm;

[0184] Path edges are conflict association relationships, which can be complaint relationships between owners and conflict events, failure relationships between equipment and conflict events, or handling relationships between maintenance personnel and owners.

[0185] The model association algorithm connects the path nodes and path edges of the conflict event dataset that has been preprocessed to establish a conflict association graph;

[0186] A3. Extract the contradiction transmission path based on the contradiction association graph;

[0187] Conduct similarity and overlap analysis on the conflict transmission paths and merge similar conflict transmission paths;

[0188] Among them, the method of similarity coincidence analysis of contradiction transmission paths is as follows:

[0189] Extract any two contradictory conduction paths as a path comparison group, extract the two longest common subpaths of the path comparison group, calculate the lengths of any two longest common subpaths, and obtain len1 and len2;

[0190] Wherein, len1 represents the length of the longest common subpath of the first contradictory conduction path in the path comparison group, and len2 represents the length of the longest common subpath of the second contradictory conduction path in the path comparison group;

[0191] The similarity of any two longest common subpaths in the path comparison group is obtained through the formula: Sim = len1 / len2;

[0192] The similarity of the two longest common subpaths is compared with a preset similarity threshold. If the similarity of the longest common subpath is higher than the preset similarity threshold, the conflicting conduction paths of the path comparison group are merged through a clustering algorithm.

[0193] The conflict conduction paths that are merged are marked as multiple conflict paths, and the conflict conduction paths that are not merged are marked as original conflict paths.

[0194] Overlay analysis module: used to extract the imbalance characteristics of property resource allocation when multiple conflicting paths are merged, and to identify whether the imbalance characteristics of property resource allocation have an overlay effect;

[0195] Among them, the method of identifying the imbalance characteristics of property resource allocation when multiple conflicting paths are merged is:

[0196] From the merging process of multiple contradictory paths, the original contradictory paths initially corresponding to the multiple contradictory paths are obtained;

[0197] When the original conflicting paths are merged into multiple conflicting paths, the numbers and merging times of the original conflicting paths are recorded during cluster merging analysis, and a path mapping table from multiple conflicting paths to original path clusters is established;

[0198] Identify the property resource imbalance characteristics that are filtered out when multiple conflicting paths are merged, and add the filtered imbalance characteristics to the path mapping table;

[0199] Extract spatial mismatch rate and time lag rate from the path mapping table as the imbalance characteristics of property resource allocation;

[0200] S1. The spatial mismatch rate is obtained as follows:

[0201] By formula: Obtain the spatial mismatch rate SMR;

[0202] Among them, ρ i is the maintenance worker density in the conflict-occurring region i in the original conflict path, that is, the ratio of the number of maintenance workers in the conflict-occurring region to the area of the region;

[0203] ρ avg The average value of the maintenance worker density in the multi-contradiction path is obtained by averaging the maintenance worker density of all relevant areas;

[0204] R i is the actual maintenance personnel coverage radius of the conflict-occurring region i in the original conflict path, R std The standard coverage radius is the theoretical coverage radius calculated based on the standard service area of a single maintenance worker set by the property.

[0205] i is the number of the area where the conflict occurs;

[0206] S2. The time lag rate is obtained as follows:

[0207] By formula: Get the time lag rate TLR;

[0208] Where M is the total number of event path work orders constructed based on active feedback and passive perception of conflicting events in the multi-contradiction path, and j is the number of the event path work order;

[0209] It should be explained that the event path work order is established based on the conflicting events and is used to assign tasks to maintenance personnel to resolve the conflicting events;

[0210] Z j It represents the standardized time deviation of the event in the original conflict path, which is calculated by the deviation ratio between the maintenance worker's maintenance time and the preset standard time T;

[0211] ΔT j The deviation value is calculated by calculating the difference between the maintenance time of the maintenance worker and the preset standard time T;

[0212] I is the indicator function, which is 1 when I meets the condition, otherwise it is 0;

[0213] Based on the imbalance characteristics of property resource allocation when multiple conflict paths are merged, determine whether the imbalance characteristics of property resource allocation of multiple conflict paths have a superposition effect;

[0214] Among them, the method to judge whether the imbalanced allocation characteristics of property resources in multiple conflict paths have a superposition effect is as follows:

[0215] By formula: Constructing a vector autoregression model to analyze the imbalance characteristics of property resource allocation in multiple conflict paths;

[0216] Among them, TLR t It represents the time lag rate at time t, reflecting the degree of deviation between the current maintenance time and the standard time;

[0217] α k TLR hysteresis effect k period t time TLR t The influence coefficient of TLR is used to measure the effect of its own historical value on the current value.

[0218] TLR t-k represents the value of TLR at time tk, that is, the historical observation value of TLR;

[0219] β k represents the effect of SMR after lag k on TLR at time t t The influence coefficient reflects the cross-effect of spatial mismatch rate on time lag rate;

[0220] SMR t-k represents the value of SMR at time tk, that is, the historical observation value of SMR;

[0221] ε 1t It represents the random error term of the TLR equation at time t, which obeys a normal distribution with a mean of 0 and captures the random factor TLR not explained by the model. t The influence of ε 2t It represents the random error term of the SMR equation at time t, which obeys a normal distribution with a mean of 0 and captures the random factor SMR not explained by the model. t the impact of;

[0222] SMR t It represents the spatial mismatch rate at time t, and quantifies the imbalance between the maintenance personnel density and coverage radius in the current area;

[0223] γ k :TLR lag k period to SMR at time t t The influence coefficient of reflects the reverse effect of time lag rate on spatial mismatch rate;

[0224] δ k :SMR lagged k period to SMR at time t t The influence coefficient of SMR describes the impact of its own historical value on the current value;

[0225] q is the lag order of the model, which determines how many past periods of TLR and SMR values are used to explain the current values. It needs to be determined based on the data characteristics and the AIC and BIC criteria of the statistical test;

[0226] t is the time, k is the number of lag periods;

[0227] is the autoregressive part of the vector autoregressive model, which respectively represents the impact of the historical values of the time lag rate and spatial mismatch rate on the current value;

[0228] If α k or δ k If the value is significantly different from zero, it indicates that the imbalance feature has temporal autocorrelation, that is, the current imbalance level is affected by the past imbalance state, which reflects the superposition effect of a single imbalance feature in the time dimension.

[0229] is the cross part in the formula, which represents the mutual influence between different imbalance characteristics; if β k or γ k If it is significantly different from zero, it indicates that different types of imbalance characteristics are correlated with each other in time.

[0230] Regional screening module: If there is a superposition effect, it is used to perform regional screening analysis on multiple conflict paths and extract the high-frequency superposition areas of all multiple conflict paths;

[0231] If there is a superposition effect, screen the multiple contradiction paths with superposition effect among all the multiple contradiction paths;

[0232] The method of extracting the high-frequency superposition area of all multiple conflict paths is:

[0233] For multiple conflict paths with superposition effects, the conflict occurrence areas of all multiple conflict paths with superposition effects are summarized;

[0234] The frequency of occurrence of each conflict region of the multi-contradiction path in different periods is calculated and compared with the preset frequency threshold. The conflict region with a frequency higher than or equal to the preset frequency threshold is marked as a high-frequency superposition region.

[0235] Resource optimization module: Based on the determined high-frequency superposition area, it is used to input the passive sensing features collected in real time into the intelligent resource allocation model to realize the intelligent prediction of unbalanced property resources and optimize the allocation of property resources;

[0236] Based on the determined high-frequency superposition area and the passive sensing features collected in real time by IoT devices;

[0237] The high-frequency superposition areas and passive perception features collected in real time by IoT devices are input into the intelligent resource allocation model to achieve intelligent prediction of unbalanced property resources and optimize the allocation of property resources.

[0238] The basic principles, main features, and advantages of the present invention are shown and described above. Those skilled in the art should understand that the present invention is not limited to the foregoing embodiments. The foregoing embodiments and descriptions are merely illustrative of the principles of the present invention. Various changes and modifications may be made to the present invention without departing from the spirit and scope of the present invention. Such changes and modifications are intended to fall within the scope of the present invention. The scope of protection claimed in the present invention is defined by the appended claims and their equivalents.

Claims

1. A property intelligent resource scheduling optimization method based on the Internet of Things, characterized by: The steps include: Obtain conflict event labels and events, as well as conflict correlations, from active feedback and passive perception by owners and property management, and construct a conflict event dataset. Construct a contradiction path model, input the contradiction event data set into the contradiction path model, extract the contradiction transmission path and conduct similarity merging analysis to obtain multiple contradiction paths; Extract the imbalance characteristics of property resource allocation when multiple conflict paths are merged, and identify whether the imbalance characteristics of property resource allocation have a superposition effect; If there is a superposition effect, conduct regional screening analysis on the multiple conflict paths and extract the high-frequency superposition areas of all the multiple conflict paths; Based on the determined high-frequency superposition area, the passive perception features collected in real time are input into the intelligent resource allocation model to realize intelligent prediction of unbalanced property resources and optimize the allocation of property resources.

2. The method for optimizing property resource scheduling based on the Internet of Things according to claim 1, characterized in that: The method of constructing the conflict event dataset is as follows: Obtain the conflict event label and the event resources and property maintenance resources in the conflict event process, and perform correlation matching to obtain the conflict correlation relationship of the conflict event; The conflict event label, event resources in the conflict event process, and property maintenance resources are used as conflict event elements; Obtain the conflict event elements and conflict correlation relationships of all conflict events between property and owners to build a conflict event dataset.

3. The method for optimizing property resource scheduling based on the Internet of Things according to claim 1, characterized in that: The contradiction path model is as follows: Perform data preprocessing on the input conflict event data set; Construct a contradiction association graph based on the contradiction event dataset after data preprocessing; The contradiction transmission path is extracted based on the contradiction association graph.

4. The method for optimizing property resource scheduling based on the Internet of Things according to claim 1, characterized in that: The method of similarity merging analysis is: Extract any two contradictory conduction paths as a path comparison group, extract the two longest common subpaths of the path comparison group, and calculate the length of any two longest common subpaths; Calculate the similarity of the lengths of any two longest common subpaths to obtain the similarity of any two longest common subpaths in the path comparison group; If the similarity of the longest common sub-path is higher than a preset similarity threshold, the contradictory conduction paths of the path comparison group are merged through a clustering algorithm.

5. The method for optimizing property resource scheduling based on the Internet of Things according to claim 1, characterized in that: The method for identifying whether the imbalance in property resource allocation has a superposition effect is: Based on the imbalance characteristics of property resource allocation when multiple conflict paths are merged, a vector autoregression model is constructed to conduct autocorrelation analysis on the imbalance characteristics of property resource allocation of multiple conflict paths. If the impact coefficient of the autocorrelation analysis is significantly different from zero, the imbalance characteristics of property resource allocation have a superimposed effect.

6. The method for optimizing property resource scheduling based on the Internet of Things according to claim 5, characterized in that: The method for obtaining the property resource allocation imbalance feature is as follows: From the merging process of multiple contradictory paths, the original contradictory paths initially corresponding to the multiple contradictory paths are obtained; Identify the property resource imbalance characteristics that are filtered out when multiple conflicting paths are merged, and add the filtered imbalance characteristics to the path mapping table; The spatial mismatch rate and time lag rate are extracted from the path mapping table as the imbalance characteristics of property resource allocation.

7. The method for optimizing property resource scheduling based on the Internet of Things according to claim 6, characterized in that: The spatial mismatch rate is obtained as follows: The spatial mismatch rate is obtained by numerically calculating the density of maintenance personnel in the conflict-occurring area in the original conflict path and the maintenance personnel density and the actual maintenance personnel coverage radius.

8. The method for optimizing property resource scheduling based on the Internet of Things according to claim 6, characterized in that: The time lag rate is obtained as follows: The time lag rate is obtained by numerically calculating the standardized time deviation value of the events in the original contradiction path.

9. The method for optimizing property resource scheduling based on the Internet of Things according to claim 7, characterized in that: The method of extracting the high-frequency superposition area of all multi-contradictory paths is: For multiple conflict paths with superposition effects, the conflict occurrence areas of all multiple conflict paths with superposition effects are summarized; The frequencies of occurrence of various conflict regions of the multi-contradiction paths in different cycles are calculated, and high-frequency superposition regions are determined based on the frequencies of occurrence of the conflict regions in different cycles.

10. A property intelligent resource scheduling optimization method based on the Internet of Things, characterized by: The method of optimizing the allocation of property resources is: Based on the determined high-frequency superposition area and the passive sensing features collected in real time by IoT devices; The high-frequency superposition areas and passive perception features collected in real time by IoT devices are input into the intelligent resource allocation model to achieve intelligent prediction of unbalanced property resources and optimize the allocation of property resources.

Images