Intelligent space internet of things management and operation system and method based on object model

Abstract

This invention discloses a smart space IoT management and operation and maintenance system and method based on a physical model, belonging to the field of data management technology. The system includes: a physical model construction module, a data acquisition and association mapping module, a fault diagnosis and analysis module, a parameter optimization and verification module, and a model iterative update module. The method includes: constructing a physical model integrating rated parameters, fault mapping relationships, and adjustable parameters; collecting data through edge nodes and establishing a dynamic association mapping with the physical model; when a fault alarm is detected, calling the physical model to perform parameter anomaly diagnosis and root cause matching; periodically optimizing equipment parameters based on operation and maintenance characteristics and verifying the effect; and finally updating the physical model to achieve iterative optimization. This invention achieves accurate fault location and autonomous parameter optimization through physical model-driven operation, forming a closed-loop management system, effectively improving operation and maintenance efficiency and system autonomy.

Description

Technical Field

[0001] This invention relates to the field of data management technology, specifically to a smart space IoT management and operation system and method based on object models. Background Technology

[0002] In the field of smart space IoT management and operation and maintenance, with the continuous increase in the types and numbers of devices, traditional operation and maintenance methods are no longer able to meet the needs of efficient and accurate management. Currently, most IoT systems analyze operation and maintenance data and device parameters separately. This fragmented processing logic leads to low operation and maintenance efficiency, inaccurate fault location, and difficulty in continuously optimizing operation and maintenance strategies.

[0003] Specifically, traditional operation and maintenance methods often rely on manual troubleshooting when facing equipment failures, which is not only time-consuming and labor-intensive but also prone to errors. Furthermore, due to the lack of effective correlation between operation and maintenance data and equipment parameters, it is impossible to extract valuable information from massive amounts of operation and maintenance data to optimize equipment parameters and operation and maintenance strategies. Equipment physical models are usually treated as static parameter templates, rarely adjusted once set, making it difficult to adapt to changes in the actual operating environment of the equipment.

[0004] Furthermore, with the rapid development of IoT technology, devices in smart spaces are becoming increasingly intelligent, placing higher demands on the real-time, accurate, and adaptive nature of operation and maintenance. Therefore, breaking down the barriers between operation and maintenance data and device parameters, and establishing an efficient and dynamic correlation analysis mechanism, has become a pressing technical problem to be solved in the field of IoT management and operation and maintenance of smart spaces. Summary of the Invention

[0005] The purpose of this invention is to provide a smart space IoT management and operation system and method based on object models to solve the problems mentioned in the background art.

[0006] To solve the above-mentioned technical problems, the present invention provides the following technical solution:

[0007] A smart space IoT management and operation method based on a physical model includes the following steps:

[0008] S1. For at least one type of target device in the smart space, construct a physical model, wherein the physical model stores the rated operating parameter range of the device, records the mapping relationship between parameter anomalies and fault characteristics generated based on historical operation and maintenance data, and sets adjustable operation and maintenance monitoring parameters and reserves an operation and maintenance data interaction interface.

[0009] S2. Based on the constructed object model, fault alarm data, real-time operation data and periodic usage data of the target device are collected through edge nodes, and the collected data is stored in the operation and maintenance database; based on the unique identifier of the object model, an association mapping between the operation and maintenance data and the corresponding object model is established;

[0010] S3. When fault alarm data of the target device is detected in the operation and maintenance database, based on the association mapping between the operation and maintenance data and the corresponding physical model, the corresponding physical model is called according to the device identifier; the rated operating parameter range stored in the physical model is extracted and compared with the real-time operating data in the operation and maintenance data to determine the parameter anomaly type; then, based on the mapping relationship between parameter anomalies and fault characteristics recorded in the physical model, the root cause of the fault and related components are output.

[0011] S4. According to the preset cycle, extract the periodic usage data and fault correlation data of the target equipment from the operation and maintenance database to generate operation and maintenance data features; input the operation and maintenance data features through the operation and maintenance data interaction interface reserved in the object model, and calculate the adjustment parameters through the preset optimization algorithm; send the adjustment parameters to the target equipment, monitor the equipment operation status to verify the adjustment effect, and if the verification meets the standard, update the operation and maintenance monitoring parameters in the object model.

[0012] S5. Synchronize the qualified and updated operation and maintenance monitoring parameters to the physical model, and combine the output root causes of failures and related component data to optimize the mapping relationship between parameter anomalies and fault characteristics in the physical model; then return to S2 and repeat the data acquisition, fault handling, parameter optimization and model update process.

[0013] Furthermore, S1 records the mapping relationship between parameter anomalies and fault characteristics generated based on historical operation and maintenance data, as detailed below:

[0014] Historical operation and maintenance data of the target equipment is collected. This historical operation and maintenance data includes abnormal equipment parameter data (such as current fluctuation, voltage deviation, and temperature drift), fault type data (such as hardware module failure and software configuration failure), operating environment data (such as ambient temperature and humidity, and power supply stability coefficient), and equipment-related data (such as the operating status of other equipment in the same communication link). The collected historical operation and maintenance data is standardized and integrated into a feature dataset D, where D = {d1, d2, ..., dn}, where n is the number of samples, and each sample di contains a parameter anomaly feature vector, a fault type label, and environmental-related features, and i ranges from 1 to n.

[0015] For each sample di in the feature dataset D, calculate the proportion p of the j-th parameter abnormal feature in the i-th sample. ij , and p ij =e xij / ∑ i∈[1,n] e xij Where xij represents the standardized value of the j-th parameter anomaly feature in the i-th sample; calculate the information entropy Ej of the j-th parameter anomaly feature, and Ej = -(1 / lnn)∑ i∈[1,n] p ij·lnp ij Calculate the weight coefficient wj of the j-th parameter anomaly feature, and wj=(1-Ej) / ∑ j∈[1,m] (1-Ej), where m is the number of parameter anomalies, ∑ j∈[1,m] wj=1; Calculate the weighted support Sw of the combination of abnormal feature parameters and fault type, and the corresponding calculation formula is:

[0016] Sw=[∑ i∈[1,n] (∑ j∈[1,k] wj·I ij )·I i target ] / n×100%,

[0017] Among them, I ij Let I be an indicator function used to identify whether sample di contains the j-th parameter anomalous feature. If sample di contains the j-th feature, then I... ij =1, otherwise 0; I i target This is an indicator function used to identify whether sample di contains the target fault type. If sample di contains the target fault type, then I... i target =1, otherwise 0; k is the number of abnormal parameter features involved in the combination; calculate the weighted confidence Cw of the combination of abnormal parameter features and the fault type, and the corresponding calculation formula is: Cw=Sw / {[(∑ i∈[1,n] ∑ j∈[1,k] wj·I ij ) / n]×100%}; Set the minimum weighted support threshold Sw_min and the minimum weighted confidence threshold Cw_min, and filter out the correlations that satisfy Sw≥Sw_min and Cw≥Cw_min, as the initial mapping relationship between parameter anomalies and fault characteristics;

[0018] An environment adaptation coefficient θ is introduced to dynamically correct the weighted confidence of the initial mapping relationship, and Cw' = Cw·θ. The environment adaptation coefficient θ is calculated based on the operating environment data, and the environment adaptation coefficient θ is equal to the ratio of the similarity between the current environment feature and the historical adapted environment feature to the standard deviation of the historical adapted environment feature. When Cw' ≥ Cw_min, the corresponding mapping relationship is retained. When Cw' < Cw_min, the corresponding mapping relationship is marked as pending update, and the next round of historical operation and maintenance data collection and association mining process is triggered to record the final effective and corrected mapping relationship into the object model.

[0019] Furthermore, in S2, based on the unique identifier of the object model, an association mapping is established between the operation and maintenance data and the corresponding object model. The specific details are as follows:

[0020] Based on the physical attributes of the target device corresponding to the thing model and the characteristics of the thing model itself, a unique thing model identifier ID_model is generated through a combined hashing algorithm; the device identifier information contained in the collected operation and maintenance data is standardized and uniformly converted into a format compatible with the unique identifier of the thing model, and the standardized device identifier is denoted as ID_device;

[0021] Calculate the association degree Score between the device identifier ID_device and the thing model identifier ID_model, where the association degree is calculated based on two dimensions: device attribute similarity and time consistency. The formula is: Score = α·Sim_attr+(1-α)·Sim_time; where Sim_attr is the device attribute similarity, which is equal to the matching degree between the SN extracted from the device identifier ID_device and the SN contained in the thing model identifier ID_model. When it is a complete match, Sim_attr = 1. When it is a partial match, it is calculated according to the proportion of matching characters. When there is no match, Sim_attr = 0; Sim_time is the time consistency, which is equal to the reciprocal of the time difference between the operation and maintenance data collection timestamp and the thing model creation timestamp, and the smaller the time difference, the closer Sim_time is to 1; α is the weight coefficient; when the association degree Score≥0.8, establish an initial association mapping between the device identifier ID_device and the thing model identifier ID_model and store it as a mapping pair <ID_device, ID_model>;

[0022] When the same device identifier ID_device corresponds to multiple thing model identifiers ID_model, calculate the historical association frequency F of each candidate thing model identifier ID_model, that is, the number of times the candidate thing model identifier ID_mode has been successfully associated with the device identifier ID_device in the past. The association degree is corrected by frequency weighting, and the corresponding calculation formula is: Score’ = Score·(1+β·F / ∑F), where β is the frequency influence factor. Select the thing model identifier ID_model with the largest Score’ as the final associated object; recalculate the association degree Score of all mapping pairs at a preset cycle. When the association degree Score<0.6, trigger the re-association process; when the thing model is updated in version, automatically migrate the original mapping pair <ID_device, old ID_model> to <ID_device, new ID_model> and retain the migration record for traceability.

[0023] Further, S3 includes the following content:

[0024] Real-time monitoring of fault alarm data D_alarm in the operation and maintenance database; when fault alarm data D_alarm is detected, extract the device identifier ID_device carried in the fault alarm data; based on the mapping pair<ID_device,ID_model> The corresponding object model identifier ID_model is retrieved by querying the object model identifier ID_model; the corresponding object model M pre-stored in the operation and maintenance database is called through the object model identifier ID_model to obtain the set R of rated operating parameter ranges and the mapping relationship table T_map of parameter anomalies and fault characteristics stored in object model M, and R={r1,r2,...,rg}, where rf represents the range interval of the f-th type of rated operating parameter, and rf=[min_rf,max_rf], where min_rf represents the minimum value of the f-th type of rated operating parameter, and max_rf represents the maximum value of the f-th type of rated operating parameter; and f takes values ​​from 1 to g, where g represents the total number of types of rated operating parameters;

[0025] Extract real-time operational data D_real associated with the device identifier ID_device from the operation and maintenance database, where D_real={dr1,dr2,...,drg}, and drf represents the real-time monitoring value of the f-th type of operational parameter; calculate the deviation δf of each type of operational parameter from the rated range, using the following formula:

[0026] When drf∈[min_rf,max_rf], the deviation δf=0;

[0027] When drf < min_rf, the deviation δf = (min_rf - drf) / min_rf;

[0028] When drf > max_rf, the deviation δf = (drf - max_rf) / max_rf;

[0029] Where δf represents the relative deviation of the f-th type of operating parameter;

[0030] Based on the abnormal feature weight coefficients wj calculated in S1, the comprehensive abnormality index δ_total is calculated, and δ_total = ∑ j∈[1,m]wj·δj, where wj is the weight coefficient of the j-th type parameter and δj is the deviation of the j-th type parameter; the parameter anomaly type is determined according to the set anomaly type threshold; for example, when δ_total∈(0,0.2], it is determined as "minor anomaly"; when δ_total∈(0.2,0.5], it is determined as "moderate anomaly"; when δ_total>0.5, it is determined as "severe anomaly"; the parameter items with δj>0 are extracted and marked as the abnormal parameter set P_abnormal, and P_abnormal={Pj|δj>0}; based on the abnormal parameter set P_abnormal, the matching parameter anomaly feature combination is retrieved in the mapping relationship table T_map of the object model M; the matching degree S_match between the retrieved matching parameter anomaly feature combination and the feature combination recorded in the mapping relationship table T_map is calculated, and the corresponding calculation formula is:

[0031] S_match=∑ j∈P_abnormal (wj·Ij_map) / ∑ j∈P_abnormal wj,

[0032] Here, Ij_map is an indicator function. If parameter Pj exists in a certain feature combination in the mapping table T_map, then Ij_map = 1; otherwise, it is 0. The environmental adaptation coefficient θ in S1 is introduced to correct the matching degree, resulting in S_match' = S_match・θ. When S_match' ≥ 0.7 (indicating high matching degree), the corresponding root cause G and associated component C in the mapping table T_map are called. When 0.5 ≤ S_match' < 0.7 (indicating medium matching degree), the root cause G and associated component C are output, and marked "Manual review recommended". When S_match' < 0.5 (indicating low matching degree), the historical operation and maintenance data supplementation collection process in S1 is triggered, and "No clear root cause has been matched yet; it is recommended to check the associated components of abnormal parameters first" is output.

[0033] The determined parameter anomaly type, fault root cause G, associated component C, and corresponding deviation data δf are integrated into a structured fault report R_fault={ID_device,T_alarm,P_abnormal,δ_total,G,C,S_match'}, where T_alarm is the alarm timestamp, stored in the operation and maintenance database, and associated with the object model M.

[0034] Furthermore, S4 includes the following:

[0035] According to a preset cycle, periodic usage data and fault correlation data of the target equipment are extracted from the operation and maintenance database. The periodic usage data of the target equipment includes the total runtime, cumulative number of start-ups and shutdowns, runtime percentage of different time periods, and average load rate within the cycle. The fault correlation data includes the number of structured fault reports generated within the cycle, the average deviation of fault correlation parameters, and fault repair time. The extracted periodic usage data and fault correlation data of the target equipment are standardized to obtain a standardized dataset D_std.

[0036] Extract the corresponding operation and maintenance data features from the standardized dataset D_std to form an operation and maintenance feature vector V, where V = [v1, v2, ..., vu], where v1 represents the operation and maintenance data feature value of the first dimension, v2 represents the operation and maintenance data feature value of the second dimension, and so on, with vu representing the operation and maintenance data feature value of the u-th dimension, and u representing the number of dimensions of the operation and maintenance feature vector V. Calculate the corresponding weight coefficient set A for the operation and maintenance feature vector V using the entropy weight method, where A = {a1, a2, ..., au}, where a1 represents the operation and maintenance data feature weight of the first dimension, a2 represents the operation and maintenance data feature weight of the second dimension, and so on, with au representing the operation and maintenance data feature weight of the u-th dimension. Based on the operation and maintenance feature vector V and the weight coefficient set A, calculate the comprehensive feature value F, where F = ∑ b∈[1,u] (ab×vb);

[0037] The comprehensive eigenvalue F is input into a preset optimization algorithm through the object model interaction interface. The adjustment parameters are calculated with the goal of minimizing the comprehensive eigenvalue F, specifically including:

[0038] The current operation and maintenance monitoring parameter Q0 in the read object model is set, the initial iteration step size is set to λ=0.05, and the termination condition is |ΔQ μ |<0.03×Q0, where ΔQ μ Let ΔQ be the adjustment amount for the μ-th iteration, where μ is the iteration number; ΔQ is the adjustment amount for the μ-th iteration. μ =λ×(F μ-1 -F μ ) / Q0, where F μ-1 F is the comprehensive eigenvalue after the μth adjustment. μ The parameter is adjusted to Q0 + ΔQ. μ-1 Post-simulation composite eigenvalues; if F μ ≥F μ-1 Update λ = λ × 0.8; after the termination condition is met, adjust the target parameter Q. t =Q0+ΔQ μ And the target adjustment parameter Q t Within the preset parameter range of the object model;

[0039] The target parameter Q is adjusted by reserving an interactive interface through the physical model.t Send the data to the target device and record the sending timestamp, the current maintenance monitoring parameter Q0, and the target adjustment parameter Q. t The comparison relationship, the association mapping pair<ID_device,ID_model> Set a verification period Tv, which is half of a preset period. Monitor the device operation data within the verification period Tv, and calculate the alarm reduction rate Ra and parameter compliance rate Rc. The alarm reduction rate Ra = (average daily alarm count in the preset period - average daily alarm count in the verification period Tv) / average daily alarm count in the preset period × 100%. The parameter compliance rate Rc = the parameter Q that meets the target adjustment within the verification period Tv. t Required runtime / total verification period Tv × 100%; when the alarm reduction rate Ra ≥ 40% and the parameter compliance rate Rc ≥ 95%, the adjustment is deemed satisfactory; if not, adjust λ = λ × 0.7, and return to recalculate the target adjustment parameter Q. t ;

[0040] If the target is met, verify the current version Vc of the object model and its associated version Vp. If Vc = Vp, directly update the operation and maintenance monitoring parameters in the object model to the target adjustment parameter Q. t Otherwise, update the model after confirming that the new model's parameter fields are compatible; if they are incompatible, create a parameter update branch and mark it as "to be adapted"; synchronously record the update time and the target adjustment parameter Q. t The alarm reduction rate Ra, parameter compliance rate Rc, and weight coefficient set A form the source tracing log.

[0041] A smart space IoT management and operation and maintenance system based on object models includes: an object model construction module, a data acquisition and association mapping module, a fault diagnosis and analysis module, a parameter optimization and verification module, and a model iteration and update module;

[0042] The object model building module builds and stores object models for target devices in the smart space. The object model includes the device's rated parameters, the mapping relationship between parameter anomalies and fault characteristics, adjustable operation and maintenance monitoring parameters, and operation and maintenance data interaction interfaces.

[0043] The data acquisition and association mapping module collects operation and maintenance data of target devices through edge nodes and stores it in the database, and establishes a dynamic association mapping between operation and maintenance data and corresponding object models based on unique identifiers;

[0044] When the fault diagnosis and analysis module detects a fault alarm, it calls the corresponding physical model to perform parameter anomaly comparison and root cause analysis, and outputs a structured fault report.

[0045] The parameter optimization and verification module periodically extracts features from operation and maintenance data, calculates and adjusts parameters through optimization algorithms, and sends them to the equipment for verification. Once the verification is successful, a parameter update instruction is generated.

[0046] The model iteration and update module updates the monitoring parameters in the physical model and optimizes the fault mapping relationship based on the verified parameters and fault analysis results, driving the system into the next operation and maintenance cycle.

[0047] Furthermore, the object model building module includes a historical data mining unit and a dynamic mapping generation unit;

[0048] The historical data mining unit standardizes and extracts features from the collected historical operation and maintenance data to form a feature dataset. The dynamic mapping generation unit calculates feature weights and association rules based on the feature dataset, and dynamically corrects and generates the mapping relationship from abnormal parameters to fault features by combining environmental factors, and stores it in the physical model.

[0049] Furthermore, the data acquisition and association mapping module includes an identification management unit and an association maintenance unit;

[0050] The identification management unit generates unique identifiers for the object model and standardizes the device identifiers in the collected data; the association maintenance unit calculates the association degree between the device identifier and the object model identifier, establishes and maintains the mapping relationship between the two, and handles many-to-many mapping conflicts and mapping migration during version updates.

[0051] Furthermore, the fault diagnosis and analysis module includes a real-time anomaly determination unit and a root cause matching unit;

[0052] The real-time anomaly detection unit monitors fault alarms in real time, extracts real-time operating data of the equipment and compares it with the rated parameters in the physical model, calculates the parameter deviation and determines the anomaly level; the root cause matching unit retrieves matching fault features in the mapping relationship of the physical model based on the set of abnormal parameters, and outputs the root cause of the fault and related component information in combination with environmental adaptability.

[0053] Furthermore, the parameter optimization and verification module includes a feature calculation unit and a closed-loop verification unit;

[0054] The feature calculation unit extracts feature vectors from periodic operation and maintenance data, calculates comprehensive feature values, and solves for the optimal equipment operation parameter adjustment values ​​through an iterative optimization algorithm; the closed-loop verification unit sends the adjustment parameters to the equipment, monitors the equipment operation status to verify the adjustment effect, and determines whether to update the object model parameters based on the preset compliance standards.

[0055] Compared with existing technologies, the beneficial effects of this invention are as follows: This invention integrates equipment rated parameters, fault mapping relationships based on historical data mining, and adjustable parameters into the physical model, so that the operation and maintenance judgment criteria are no longer statically preset, but can be dynamically adjusted and optimized based on actual operating data through preset algorithms. This design enables the physical model to have self-evolution capabilities, significantly improving the system's adaptability to complex and ever-changing operating environments. By calculating a multi-parameter weighted comprehensive anomaly index and combining it with an environmental adaptation coefficient to correct the fault feature matching degree, this invention effectively reduces false alarms caused by fluctuations in a single parameter, can more accurately identify key anomaly combinations, and output fault root causes and related components with high confidence, providing a clear and reliable basis for maintenance decisions and greatly improving troubleshooting efficiency. This invention uses fault diagnosis results as input for parameter optimization, automatically optimizes through iterative algorithms, and sends them to the equipment for verification, forming a complete self-regulating loop. This system can not only passively respond to faults, but also actively adjust equipment operating parameters to eliminate hidden dangers, thereby effectively reducing the fault incidence rate and extending equipment life. By automating and intelligentizing the entire process, this invention frees maintenance personnel from tedious monitoring, rule configuration, and troubleshooting, reduces reliance on specific experts and the risk of human error, and achieves low-cost, high-efficiency, and standardized operation and maintenance of large-scale IoT facilities. Attached Figure Description

[0056] The accompanying drawings are provided to further illustrate the invention and form part of the specification. They are used in conjunction with embodiments of the invention to explain the invention and do not constitute a limitation thereof. In the drawings:

[0057] Figure 1 This is a schematic diagram of a smart space IoT management and operation system based on a physical model, according to the present invention. Detailed Implementation

[0058] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0059] Please see Figure 1 The present invention provides the following technical solution:

[0060] A smart space IoT management and operation and maintenance system based on object models includes: an object model construction module, a data acquisition and association mapping module, a fault diagnosis and analysis module, a parameter optimization and verification module, and a model iteration and update module;

[0061] The object model building module builds and stores object models for target devices in the smart space. The object model includes the device's rated parameters, the mapping relationship between parameter anomalies and fault characteristics, adjustable operation and maintenance monitoring parameters, and operation and maintenance data interaction interfaces.

[0062] The data acquisition and association mapping module collects operation and maintenance data of target devices through edge nodes and stores it in the database, and establishes a dynamic association mapping between operation and maintenance data and corresponding object models based on unique identifiers;

[0063] When the fault diagnosis and analysis module detects a fault alarm, it calls the corresponding physical model to perform parameter anomaly comparison and root cause analysis, and outputs a structured fault report.

[0064] The parameter optimization and verification module periodically extracts features from operation and maintenance data, calculates and adjusts parameters through optimization algorithms, and sends them to the equipment for verification. Once the verification is successful, a parameter update instruction is generated.

[0065] The model iteration and update module updates the monitoring parameters in the physical model and optimizes the fault mapping relationship based on the verified parameters and fault analysis results, driving the system into the next operation and maintenance cycle.

[0066] The object model building module includes a historical data mining unit and a dynamic mapping generation unit;

[0067] The historical data mining unit standardizes and extracts features from the collected historical operation and maintenance data to form a feature dataset. The dynamic mapping generation unit calculates feature weights and association rules based on the feature dataset, and dynamically corrects and generates the mapping relationship from abnormal parameters to fault features by combining environmental factors, and stores it in the physical model.

[0068] The data acquisition and association mapping module includes an identification management unit and an association maintenance unit;

[0069] The identification management unit generates unique identifiers for the object model and standardizes the device identifiers in the collected data; the association maintenance unit calculates the association degree between the device identifier and the object model identifier, establishes and maintains the mapping relationship between the two, and handles many-to-many mapping conflicts and mapping migration during version updates.

[0070] The fault diagnosis and analysis module includes a real-time anomaly detection unit and a root cause matching unit;

[0071] The real-time anomaly detection unit monitors fault alarms in real time, extracts real-time operating data of the equipment and compares it with the rated parameters in the physical model, calculates the parameter deviation and determines the anomaly level; the root cause matching unit retrieves matching fault features in the mapping relationship of the physical model based on the set of abnormal parameters, and outputs the root cause of the fault and related component information in combination with environmental adaptability.

[0072] The parameter optimization and verification module includes a feature calculation unit and a closed-loop verification unit;

[0073] The feature calculation unit extracts feature vectors from periodic operation and maintenance data, calculates comprehensive feature values, and solves for the optimal equipment operation parameter adjustment values ​​through an iterative optimization algorithm; the closed-loop verification unit sends the adjustment parameters to the equipment, monitors the equipment operation status to verify the adjustment effect, and determines whether to update the object model parameters based on the preset compliance standards.

[0074] A smart space IoT management and operation method based on a physical model includes the following steps:

[0075] S1. For at least one type of target device in the smart space, construct a physical model, wherein the physical model stores the rated operating parameter range of the device, records the mapping relationship between parameter anomalies and fault characteristics generated based on historical operation and maintenance data, and sets adjustable operation and maintenance monitoring parameters and reserves an operation and maintenance data interaction interface.

[0076] S2. Based on the constructed object model, fault alarm data, real-time operation data and periodic usage data of the target device are collected through edge nodes, and the collected data is stored in the operation and maintenance database; based on the unique identifier of the object model, an association mapping between the operation and maintenance data and the corresponding object model is established;

[0077] S3. When fault alarm data of the target device is detected in the operation and maintenance database, based on the association mapping between the operation and maintenance data and the corresponding physical model, the corresponding physical model is called according to the device identifier; the rated operating parameter range stored in the physical model is extracted and compared with the real-time operating data in the operation and maintenance data to determine the parameter anomaly type; then, based on the mapping relationship between parameter anomalies and fault characteristics recorded in the physical model, the root cause of the fault and related components are output.

[0078] S4. According to the preset cycle, extract the periodic usage data and fault correlation data of the target equipment from the operation and maintenance database to generate operation and maintenance data features; input the operation and maintenance data features through the operation and maintenance data interaction interface reserved in the object model, and calculate the adjustment parameters through the preset optimization algorithm; send the adjustment parameters to the target equipment, monitor the equipment operation status to verify the adjustment effect, and if the verification meets the standard, update the operation and maintenance monitoring parameters in the object model.

[0079] S5. Synchronize the qualified and updated operation and maintenance monitoring parameters to the physical model, and combine the output root causes of failures and related component data to optimize the mapping relationship between parameter anomalies and fault characteristics in the physical model; then return to S2 and repeat the data acquisition, fault handling, parameter optimization and model update process.

[0080] S1 records the mapping relationship between parameter anomalies and fault characteristics generated based on historical operation and maintenance data. The specific content is as follows:

[0081] Historical operation and maintenance data of the target equipment is collected. This historical operation and maintenance data includes abnormal equipment parameter data (such as current fluctuation, voltage deviation, and temperature drift), fault type data (such as hardware module failure and software configuration failure), operating environment data (such as ambient temperature and humidity, and power supply stability coefficient), and equipment-related data (such as the operating status of other equipment in the same communication link). The collected historical operation and maintenance data is standardized and integrated into a feature dataset D, where D = {d1, d2, ..., dn}, where n is the number of samples, and each sample di contains a parameter anomaly feature vector, a fault type label, and environmental-related features, and i ranges from 1 to n.

[0082] For each sample di in the feature dataset D, calculate the proportion p of the j-th parameter abnormal feature in the i-th sample. ij , and p ij =e xij / ∑ i∈[1,n] e xij Where xij represents the standardized value of the j-th parameter anomaly feature in the i-th sample; calculate the information entropy Ej of the j-th parameter anomaly feature, and Ej = -(1 / lnn)∑ i∈[1,n] p ij ·lnp ij Calculate the weight coefficient wj of the j-th parameter anomaly feature, and wj=(1-Ej) / ∑ j∈[1,m] (1-Ej), where m is the number of parameter anomalies, ∑ j∈[1,m] wj=1; Calculate the weighted support Sw of the combination of abnormal feature parameters and fault type, and the corresponding calculation formula is:

[0083] Sw=[∑ i∈[1,n] (∑ j∈[1,k] wj·I ij )·I i target ] / n×100%,

[0084] Among them, I ij Let I be an indicator function used to identify whether sample di contains the j-th parameter anomalous feature. If sample di contains the j-th feature, then I... ij =1, otherwise 0; I i target This is an indicator function used to identify whether sample di contains the target fault type. If sample di contains the target fault type, then I... i target =1, otherwise 0; k is the number of abnormal parameter features involved in the combination; calculate the weighted confidence Cw of the combination of abnormal parameter features and the fault type, and the corresponding calculation formula is: Cw=Sw / {[(∑ i∈[1,n] ∑ j∈[1,k]wj·I ij ) / n]×100%}; Set the minimum weighted support threshold Sw_min and the minimum weighted confidence threshold Cw_min, and filter out the correlations that satisfy Sw≥Sw_min and Cw≥Cw_min, as the initial mapping relationship between parameter anomalies and fault characteristics;

[0085] An environment adaptation coefficient θ is introduced to dynamically correct the weighted confidence of the initial mapping relationship, and Cw' = Cw·θ. The environment adaptation coefficient θ is calculated based on the operating environment data, and the environment adaptation coefficient θ is equal to the ratio of the similarity between the current environment feature and the historical adapted environment feature to the standard deviation of the historical adapted environment feature. When Cw' ≥ Cw_min, the corresponding mapping relationship is retained. When Cw' < Cw_min, the corresponding mapping relationship is marked as pending update, and the next round of historical operation and maintenance data collection and association mining process is triggered to record the final effective and corrected mapping relationship into the object model.

[0086] In S2, a mapping between operation and maintenance data and the corresponding object model is established based on the unique identifier of the object model. The specific details are as follows:

[0087] Based on the physical attributes of the target device corresponding to the object model and the characteristics of the object model itself, a unique object model identifier ID_model is generated by combining hash algorithms; the device identification information contained in the collected operation and maintenance data is standardized and uniformly converted into a format compatible with the unique identifier of the object model, and the standardized device identifier is recorded as ID_device;

[0088] In this embodiment, the physical attributes of the target device corresponding to the object model, such as the device hardware serial number SN and manufacturer code M; the object model's own characteristics, such as the model version number V and creation timestamp T; a unique identifier ID_model is generated by a combined hash algorithm, with the formula: ID_model=Hash(SN+M+V+T), where Hash is the SHA-256 hash function, ensuring that the ID_model of different object models is unique and irreversible; the device identification information contained in the collected operation and maintenance data, such as the device name, communication address, and temporary number reported by the edge node, is standardized and uniformly converted into a format compatible with the unique identifier of the object model. Specifically: if the operation and maintenance data contains the device hardware serial number SN, the SN is directly extracted as the basic identifier; if the operation and maintenance data does not contain SN, the SN is obtained by matching the device communication address with a preset address-SN mapping table; the standardized device identifier is recorded as ID_device.

[0089] Calculate the correlation degree Score between the computing device identifier ID_device and the physical model identifier ID_model. The correlation degree is calculated based on two dimensions: device attribute similarity and time consistency. The formula is: Score = α·Sim_attr+(1-α)·Sim_time; where Sim_attr is the device attribute similarity, which is equal to the matching degree between the SN extracted from the device identifier ID_device and the SN included in the physical model identifier ID_model. When there is a complete match, Sim_attr = 1. When there is a partial match, it is calculated according to the proportion of the matching characters. When there is no match, Sim_attr = 0; Sim_time is the time consistency, which is equal to the reciprocal of the time difference between the operation and maintenance data collection timestamp and the physical model creation timestamp. The smaller the time difference, the closer Sim_time is to 1; α is the weight coefficient; when the correlation degree Score≥0.8, establish the initial association mapping between the device identifier ID_device and the physical model identifier ID_model, and store it as a mapping pair <ID_device, ID_model>;

[0090] When the same device identifier ID_device corresponds to multiple physical model identifiers ID_model, calculate the historical association frequency F of each candidate physical model identifier ID_model, that is, the number of times the candidate physical model identifier ID_mode has been successfully associated with the device identifier ID_device in the past. The correlation degree is corrected by frequency weighting. The corresponding calculation formula is: Score’ = Score·(1+β·F / ∑F), where β is the frequency influence factor. Select the physical model identifier ID_model with the largest Score’ as the final associated object; Recalculate the correlation degree Score of all mapping pairs at a preset cycle. When the correlation degree Score<0.6, trigger the re-association process; when the physical model is updated, automatically migrate the original mapping pair <ID_device, old ID_model> to <ID_device, new ID_model>, and retain the migration record for traceability.

[0091] S3 includes the following:

[0092] Real-time monitoring of fault alarm data D_alarm in the operation and maintenance database; when fault alarm data D_alarm is detected, extract the device identifier ID_device carried in the fault alarm data; based on the mapping pair<ID_device,ID_model> The corresponding object model identifier ID_model is retrieved by querying the object model identifier ID_model; the corresponding object model M pre-stored in the operation and maintenance database is called through the object model identifier ID_model to obtain the set R of rated operating parameter ranges and the mapping relationship table T_map of parameter anomalies and fault characteristics stored in object model M, and R={r1,r2,...,rg}, where rf represents the range interval of the f-th type of rated operating parameter, and rf=[min_rf,max_rf], where min_rf represents the minimum value of the f-th type of rated operating parameter, and max_rf represents the maximum value of the f-th type of rated operating parameter; and f takes values ​​from 1 to g, where g represents the total number of types of rated operating parameters;

[0093] Extract real-time operational data D_real associated with the device identifier ID_device from the operation and maintenance database, where D_real={dr1,dr2,...,drg}, and drf represents the real-time monitoring value of the f-th type of operational parameter; calculate the deviation δf of each type of operational parameter from the rated range, using the following formula:

[0094] When drf∈[min_rf,max_rf], the deviation δf=0;

[0095] When drf < min_rf, the deviation δf = (min_rf - drf) / min_rf;

[0096] When drf > max_rf, the deviation δf = (drf - max_rf) / max_rf;

[0097] Where δf represents the relative deviation of the f-th type of operating parameter;

[0098] Based on the abnormal feature weight coefficients wj calculated in S1, the comprehensive abnormality index δ_total is calculated, and δ_total = ∑ j∈[1,m]wj·δj, where wj is the weight coefficient of the j-th type parameter and δj is the deviation of the j-th type parameter; the parameter anomaly type is determined according to the set anomaly type threshold; for example, when δ_total∈(0,0.2], it is determined as "minor anomaly"; when δ_total∈(0.2,0.5], it is determined as "moderate anomaly"; when δ_total>0.5, it is determined as "severe anomaly"; the parameter items with δj>0 are extracted and marked as the abnormal parameter set P_abnormal, and P_abnormal={Pj|δj>0}; based on the abnormal parameter set P_abnormal, the matching parameter anomaly feature combination is retrieved in the mapping relationship table T_map of the object model M; the matching degree S_match between the retrieved matching parameter anomaly feature combination and the feature combination recorded in the mapping relationship table T_map is calculated, and the corresponding calculation formula is:

[0099] S_match=∑ j∈P_abnormal (wj·Ij_map) / ∑ j∈P_abnormal wj,

[0100] Here, Ij_map is an indicator function. If parameter Pj exists in a certain feature combination in the mapping table T_map, then Ij_map = 1; otherwise, it is 0. The environmental adaptation coefficient θ in S1 is introduced to correct the matching degree, resulting in S_match' = S_match・θ. When S_match' ≥ 0.7 (indicating high matching degree), the corresponding root cause G and associated component C in the mapping table T_map are called. When 0.5 ≤ S_match' < 0.7 (indicating medium matching degree), the root cause G and associated component C are output, and marked "Manual review recommended". When S_match' < 0.5 (indicating low matching degree), the historical operation and maintenance data supplementation collection process in S1 is triggered, and "No clear root cause has been matched yet; it is recommended to check the associated components of abnormal parameters first" is output.

[0101] The determined parameter anomaly type, fault root cause G, associated component C, and corresponding deviation data δf are integrated into a structured fault report R_fault={ID_device,T_alarm,P_abnormal,δ_total,G,C,S_match'}, where T_alarm is the alarm timestamp, stored in the operation and maintenance database, and associated with the object model M.

[0102] S4 includes the following:

[0103] According to a preset cycle, periodic usage data and fault correlation data of the target equipment are extracted from the operation and maintenance database. The periodic usage data of the target equipment includes the total runtime, cumulative number of start-ups and shutdowns, runtime percentage of different time periods, and average load rate within the cycle. The fault correlation data includes the number of structured fault reports generated within the cycle, the average deviation of fault correlation parameters, and fault repair time. The extracted periodic usage data and fault correlation data of the target equipment are standardized to obtain a standardized dataset D_std.

[0104] Extract the corresponding operation and maintenance data features from the standardized dataset D_std to form an operation and maintenance feature vector V, where V = [v1, v2, ..., vu], where v1 represents the operation and maintenance data feature value of the first dimension, v2 represents the operation and maintenance data feature value of the second dimension, and so on, with vu representing the operation and maintenance data feature value of the u-th dimension, and u representing the number of dimensions of the operation and maintenance feature vector V. Calculate the corresponding weight coefficient set A for the operation and maintenance feature vector V using the entropy weight method, where A = {a1, a2, ..., au}, where a1 represents the operation and maintenance data feature weight of the first dimension, a2 represents the operation and maintenance data feature weight of the second dimension, and so on, with au representing the operation and maintenance data feature weight of the u-th dimension. Based on the operation and maintenance feature vector V and the weight coefficient set A, calculate the comprehensive feature value F, where F = ∑ b∈[1,u] (ab×vb);

[0105] The comprehensive eigenvalue F is input into a preset optimization algorithm through the object model interaction interface. The adjustment parameters are calculated with the goal of minimizing the comprehensive eigenvalue F, specifically including:

[0106] The current operation and maintenance monitoring parameter Q0 in the read object model is set, the initial iteration step size is set to λ=0.05, and the termination condition is |ΔQ μ |<0.03×Q0, where ΔQ μ Let ΔQ be the adjustment amount for the μ-th iteration, where μ is the iteration number; ΔQ is the adjustment amount for the μ-th iteration. μ =λ×(F μ-1 -F μ ) / Q0, where F μ-1 F is the comprehensive eigenvalue after the μth adjustment. μ The parameter is adjusted to Q0 + ΔQ. μ-1 Post-simulation composite eigenvalues; if F μ ≥F μ-1 Update λ = λ × 0.8; after the termination condition is met, adjust the target parameter Q. t =Q0+ΔQ μ And the target adjustment parameter Q t Within the preset parameter range of the object model;

[0107] The target parameter Q is adjusted by reserving an interactive interface through the physical model.t Send the data to the target device and record the sending timestamp, the current maintenance monitoring parameter Q0, and the target adjustment parameter Q. t The comparison relationship, the association mapping pair<ID_device,ID_model> Set a verification period Tv, which is half of a preset period. Monitor the device operation data within the verification period Tv, and calculate the alarm reduction rate Ra and parameter compliance rate Rc. The alarm reduction rate Ra = (average daily alarm count in the preset period - average daily alarm count in the verification period Tv) / average daily alarm count in the preset period × 100%. The parameter compliance rate Rc = the parameter Q that meets the target adjustment within the verification period Tv. t Required runtime / total verification period Tv × 100%; when the alarm reduction rate Ra ≥ 40% and the parameter compliance rate Rc ≥ 95%, the adjustment is deemed satisfactory; if not, adjust λ = λ × 0.7, and return to recalculate the target adjustment parameter Q. t ;

[0108] If the target is met, verify the current version Vc of the object model and its associated version Vp. If Vc = Vp, directly update the operation and maintenance monitoring parameters in the object model to the target adjustment parameter Q. t Otherwise, update the model after confirming that the new model's parameter fields are compatible; if they are incompatible, create a parameter update branch and mark it as "to be adapted"; synchronously record the update time and the target adjustment parameter Q. t The alarm reduction rate Ra, parameter compliance rate Rc, and weight coefficient set A form the source tracing log.

[0109] It should be noted that, in this document, relational terms such as "first" and "second" are used only to distinguish one entity or operation from another, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Furthermore, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such process, method, article, or apparatus.

[0110] Finally, it should be noted that the above descriptions are merely preferred embodiments 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 foregoing embodiments, those skilled in the art can still modify the technical solutions described in the foregoing embodiments or make equivalent substitutions for some of the technical features. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the protection scope of the present invention.

Claims

1. A smart space IoT management and operation method based on a physical model, characterized in that: The method includes the following steps: S1. For at least one type of target device in the smart space, construct a physical model, wherein the physical model stores the rated operating parameter range of the device, records the mapping relationship between parameter anomalies and fault characteristics generated based on historical operation and maintenance data, and sets adjustable operation and maintenance monitoring parameters and reserves an operation and maintenance data interaction interface. S2. Based on the constructed object model, fault alarm data, real-time operation data and periodic usage data of the target device are collected through edge nodes, and the collected data is stored in the operation and maintenance database; based on the unique identifier of the object model, an association mapping between the operation and maintenance data and the corresponding object model is established; S3. When fault alarm data of the target device is detected in the operation and maintenance database, based on the association mapping between the operation and maintenance data and the corresponding physical model, the corresponding physical model is called according to the device identifier; the rated operating parameter range stored in the physical model is extracted and compared with the real-time operating data in the operation and maintenance data to determine the parameter anomaly type; then, based on the mapping relationship between parameter anomalies and fault characteristics recorded in the physical model, the root cause of the fault and related components are output. S4. According to the preset cycle, extract the periodic usage data and fault correlation data of the target equipment from the operation and maintenance database to generate operation and maintenance data features; input the operation and maintenance data features through the operation and maintenance data interaction interface reserved in the object model, and calculate the adjustment parameters through the preset optimization algorithm; send the adjustment parameters to the target equipment, monitor the equipment operation status to verify the adjustment effect, and if the verification meets the standard, update the operation and maintenance monitoring parameters in the object model. S5. Synchronize the qualified and updated operation and maintenance monitoring parameters to the physical model, and combine the output root causes of failures and related component data to optimize the mapping relationship between parameter anomalies and fault characteristics in the physical model; then return to S2 and repeat the data acquisition, fault handling, parameter optimization and model update process.

2. The method for intelligent space IoT management and operation and maintenance based on a thing model according to claim 1, characterized in that: S1 records the mapping relationship between parameter anomalies and fault characteristics generated based on historical operation and maintenance data, as detailed below: Historical operation and maintenance data of the target equipment are collected. The historical operation and maintenance data includes abnormal equipment parameter data, fault type data, operating environment data, and equipment-related data. The collected historical operation and maintenance data are standardized and integrated into a feature dataset D, where D={d1,d2,...,dn}, where n is the number of samples, and each sample di contains a parameter anomaly feature vector, a fault type label, and an environment-related feature, and i takes values ​​from 1 to n. For each sample di in the feature dataset D, calculate the proportion p of the j-th parameter abnormal feature in the i-th sample. ij , and p ij =e xij / ∑ i∈[1,n] e xij Where xij represents the standardized value of the j-th parameter anomaly feature in the i-th sample; calculate the information entropy Ej of the j-th parameter anomaly feature, and Ej = -(1 / lnn)∑ i∈[1,n] p ij ·lnp ij Calculate the weight coefficient wj of the j-th parameter anomaly feature, and wj=(1-Ej) / ∑ j∈[1,m] (1-Ej), where m is the number of parameter anomalies, ∑ j∈[1,m] wj=1; Calculate the weighted support Sw of the combination of abnormal feature parameters and fault type, and the corresponding calculation formula is: Sw=[∑ i∈[1,n] (∑ j∈[1,k] wj·I ij )·I i target ] / n×100%, Among them, I ij Let I be an indicator function used to identify whether sample di contains the j-th parameter anomalous feature. If sample di contains the j-th feature, then I... ij =1, otherwise 0; I i target This is an indicator function used to identify whether sample di contains the target fault type. If sample di contains the target fault type, then I... i target =1, otherwise 0; k is the number of abnormal features of the parameters involved in the combination; The weighted confidence score Cw for the combination of abnormal parameter features and fault type is calculated using the following formula: Cw = Sw / {[(∑ i∈[1,n] ∑ j∈[1,k] wj·I ij ) / n]×100%}; Set the minimum weighted support threshold Sw_min and the minimum weighted confidence threshold Cw_min, and filter out the correlations that satisfy Sw≥Sw_min and Cw≥Cw_min, as the initial mapping relationship between parameter anomalies and fault characteristics; An environment adaptation coefficient θ is introduced to dynamically correct the weighted confidence of the initial mapping relationship, and Cw' = Cw·θ. The environment adaptation coefficient θ is calculated based on the operating environment data, and the environment adaptation coefficient θ is equal to the ratio of the similarity between the current environment feature and the historical adapted environment feature to the standard deviation of the historical adapted environment feature. When Cw' ≥ Cw_min, the corresponding mapping relationship is retained. When Cw' < Cw_min, the corresponding mapping relationship is marked as pending update, and the next round of historical operation and maintenance data collection and association mining process is triggered to record the final effective and corrected mapping relationship into the object model.

3. The method for intelligent space IoT management and operation based on a thing model according to claim 2, characterized in that: In S2, based on the unique identifier of the object model, an association mapping between operation and maintenance data and the corresponding object model is established, and the specific content is as follows: Based on the physical attributes of the target device corresponding to the object model and the characteristics of the object model itself, a unique object model identifier ID_model is generated through a combined hash algorithm; the device identifier information included in the collected operation and maintenance data is standardized and uniformly converted into a format compatible with the unique identifier of the object model, and the standardized device identifier is denoted as ID_device; Calculate the association degree Score between the device identifier ID_device and the object model identifier ID_model. The association degree is calculated based on two dimensions: device attribute similarity and time consistency. The formula is: Score = α·Sim_attr + (1 - α)·Sim_time; where Sim_attr is the device attribute similarity, α is the weight coefficient, Sim_time is the time consistency, which is equal to the reciprocal of the time difference between the operation and maintenance data collection timestamp and the object model creation timestamp; when the association degree Score ≥ 0.8, an initial association mapping between the device identifier ID_device and the object model identifier ID_model is established and stored as a mapping pair <ID_device, ID_model>; When the same device identifier ID_device corresponds to multiple object model identifiers ID_model, calculate the historical association frequency F of each candidate object model identifier ID_model, that is, the number of times the candidate object model identifier ID_mode has been successfully associated with the device identifier ID_device in the past. The association degree is corrected by frequency weighting, and the corresponding calculation formula is: Score’ = Score·(1 + β·F / ∑F), where β is the frequency influence factor, and the object model identifier ID_model with the largest Score’ is selected as the final association object; recalculate the association degree Score of all mapping pairs at a preset cycle. When the association degree Score < 0.6, trigger the re - association process; When the object model is updated in version, automatically migrate the original mapping pair <ID_device, old ID_model> to <ID_device, new ID_model>, and retain the migration record for traceability.

4. The method for intelligent space IoT management and operation and maintenance based on a thing model according to claim 3, characterized in that: S3 includes the following content: Real-time monitoring of fault alarm data D_alarm in the operation and maintenance database; when fault alarm data D_alarm is detected, extract the device identifier ID_device carried in the fault alarm data; based on the mapping pair<ID_device,ID_model> The corresponding object model identifier ID_model is retrieved by querying the object model identifier ID_model; the corresponding object model M pre-stored in the operation and maintenance database is called through the object model identifier ID_model to obtain the set R of rated operating parameter ranges and the mapping relationship table T_map of parameter anomalies and fault characteristics stored in object model M, and R={r1,r2,...,rg}, where rf represents the range interval of the f-th type of rated operating parameter, and rf=[min_rf,max_rf], where min_rf represents the minimum value of the f-th type of rated operating parameter, and max_rf represents the maximum value of the f-th type of rated operating parameter; and f takes values ​​from 1 to g, where g represents the total number of types of rated operating parameters; Extract real-time operational data D_real associated with the device identifier ID_device from the operation and maintenance database, where D_real={dr1,dr2,...,drg}, and drf represents the real-time monitoring value of the f-th type of operational parameter; calculate the deviation δf of each type of operational parameter from the rated range, using the following formula: When drf∈[min_rf,max_rf], the deviation δf=0; When drf < min_rf, the deviation δf = (min_rf - drf) / min_rf; When drf > max_rf, the deviation δf = (drf - max_rf) / max_rf; Where δf represents the relative deviation of the f-th type of operating parameter; Based on the abnormal feature weight coefficients wj calculated in S1, the comprehensive abnormality index δ_total is calculated, and δ_total = ∑ j∈[1,m] wj·δj, where wj is the weight coefficient of the j-th type parameter and δj is the deviation of the j-th type parameter; the parameter anomaly type is determined according to the set anomaly type threshold; the parameter items with δj>0 are extracted and marked as the anomalous parameter set P_abnormal, and P_abnormal={Pj|δj>0}; based on the anomalous parameter set P_abnormal, matching parameter anomaly feature combinations are retrieved from the mapping relationship table T_map of the object model M; the matching degree S_match between the retrieved matching parameter anomaly feature combinations and the feature combinations recorded in the mapping relationship table T_map is calculated, and the corresponding calculation formula is: S_match=∑ j∈P_abnormal (wj·Ij_map) / ∑ j∈P_abnormal wj, Here, Ij_map is an indicator function. If parameter Pj exists in a certain feature combination in the mapping table T_map, then Ij_map=1; otherwise, it is 0. The environmental adaptation coefficient θ in S1 is introduced to correct the matching degree, resulting in S_match'=S_match・θ. When S_match'≥0.7, the corresponding root cause G and associated component C in the mapping table T_map are called. When 0.5≤S_match'<0.7, the root cause G and associated component C are output, and marked "Manual review recommended". When S_match'<0.5, the historical operation and maintenance data supplementation collection process in S1 is triggered, and "No clear root cause has been matched yet. It is recommended to check the associated components of abnormal parameters first" is output. The determined parameter anomaly type, fault root cause G, associated component C, and corresponding deviation data δf are integrated into a structured fault report R_fault={ID_device,T_alarm,P_abnormal,δ_total,G,C,S_match'}, where T_alarm is the alarm timestamp, stored in the operation and maintenance database, and associated with the object model M.

5. The method for intelligent space IoT management and operation and maintenance based on a thing model according to claim 4, characterized in that: S4 includes the following: According to a preset cycle, periodic usage data and fault-related data of the target equipment are extracted from the operation and maintenance database. The periodic usage data of the target equipment includes the total runtime of the equipment within the cycle, the cumulative number of start-stops, the percentage of runtime in different time periods, and the average load rate. The fault-related data includes the number of structured fault reports generated within the cycle, the average deviation of fault-related parameters, and the fault repair time. The extracted periodic usage data and fault correlation data of the target equipment are standardized to obtain a standardized dataset D_std; Extract the corresponding operation and maintenance data features from the standardized dataset D_std to form an operation and maintenance feature vector V, where V = [v1, v2, ..., vu], where v1 represents the operation and maintenance data feature value of the first dimension, v2 represents the operation and maintenance data feature value of the second dimension, and so on, with vu representing the operation and maintenance data feature value of the u-th dimension, and u representing the number of dimensions of the operation and maintenance feature vector V. Calculate the corresponding weight coefficient set A for the operation and maintenance feature vector V using the entropy weight method, where A = {a1, a2, ..., au}, where a1 represents the operation and maintenance data feature weight of the first dimension, a2 represents the operation and maintenance data feature weight of the second dimension, and so on, with au representing the operation and maintenance data feature weight of the u-th dimension. Based on the operation and maintenance feature vector V and the weight coefficient set A, calculate the comprehensive feature value F, where F = ∑ b∈[1,u] (ab×vb); The comprehensive eigenvalue F is input into a preset optimization algorithm through the object model interaction interface. The adjustment parameters are calculated with the goal of minimizing the comprehensive eigenvalue F, specifically including: The current operation and maintenance monitoring parameter Q0 in the read object model is set, the initial iteration step size is set to λ=0.05, and the termination condition is |ΔQ μ |<0.03×Q0, where ΔQ μ Let ΔQ be the adjustment amount for the μ-th iteration, where μ is the iteration number; ΔQ is the adjustment amount for the μ-th iteration. μ =λ×(F μ-1 -F μ ) / Q0, where F μ-1 F is the comprehensive eigenvalue after the μth adjustment. μ The parameter is adjusted to Q0 + ΔQ. μ-1 Post-simulation composite eigenvalues; if F μ ≥F μ-1 Update λ = λ × 0.8; after the termination condition is met, adjust the target parameter Q. t =Q0+ΔQ μ And the target adjustment parameter Q t Within the preset parameter range of the object model; The target parameter Q is adjusted by reserving an interactive interface through the physical model. t Send the data to the target device and record the sending timestamp, the current maintenance monitoring parameter Q0, and the target adjustment parameter Q. t The comparison relationship, the association mapping pair<ID_device,ID_model> Set a verification period Tv, which is half of a preset period. Monitor the device operation data within the verification period Tv, and calculate the alarm reduction rate Ra and parameter compliance rate Rc. The alarm reduction rate Ra = (average daily alarm count in the preset period - average daily alarm count in the verification period Tv) / average daily alarm count in the preset period × 100%. The parameter compliance rate Rc = the parameter Q that meets the target adjustment within the verification period Tv. t Required runtime / total verification period Tv × 100%; when the alarm reduction rate Ra ≥ 40% and the parameter compliance rate Rc ≥ 95%, the adjustment is deemed satisfactory; if not, adjust λ = λ × 0.7, and return to recalculate the target adjustment parameter Q. t ; If the target is met, verify the current version Vc of the object model and its associated version Vp. If Vc = Vp, directly update the operation and maintenance monitoring parameters in the object model to the target adjustment parameter Q. t Otherwise, update the model after confirming that the new model's parameter fields are compatible; if they are incompatible, create a parameter update branch and mark it as "to be adapted"; synchronously record the update time and the target adjustment parameter Q. t The alarm reduction rate Ra, parameter compliance rate Rc, and weight coefficient set A form the source tracing log.

6. A smart space IoT management and operation system based on a thing model, applied to the smart space IoT management and operation method based on a thing model as described in any one of claims 1-5, characterized in that: The system includes: a physical model construction module, a data acquisition and association mapping module, a fault diagnosis and analysis module, a parameter optimization and verification module, and a model iteration and update module; The object model building module constructs and stores object models for target devices within the smart space. The object model includes device rated parameters, mapping relationships between parameter anomalies and fault characteristics, adjustable operation and maintenance monitoring parameters, and operation and maintenance data interaction interfaces. The data acquisition and association mapping module collects the operation and maintenance data of the target device through edge nodes and stores it in the database, and establishes a dynamic association mapping between the operation and maintenance data and the corresponding object model based on the unique identifier; When the fault diagnosis and analysis module detects a fault alarm, it calls the corresponding physical model to perform parameter anomaly comparison and root cause analysis, and outputs a structured fault report. The parameter optimization and verification module periodically extracts operation and maintenance data features, calculates and adjusts parameters through optimization algorithms, and sends them to the equipment for verification. After verification, it generates a parameter update instruction. The model iteration and update module updates the monitoring parameters in the physical model and optimizes the fault mapping relationship based on the verified parameters and fault analysis results, driving the system into the next operation and maintenance cycle.

7. The intelligent space IoT management and operation system based on a thing model according to claim 6, characterized in that: The object model construction module includes a historical data mining unit and a dynamic mapping generation unit; The historical data mining unit performs standardized processing and feature extraction on the collected historical operation and maintenance data to form a feature dataset. The dynamic mapping generation unit calculates feature weights and association rules based on the feature dataset, and combines environmental factors to dynamically correct and generate the mapping relationship from abnormal parameters to fault features, and stores it in the physical model.

8. The intelligent space IoT management and operation system based on a thing model according to claim 6, characterized in that: The data acquisition and association mapping module includes an identification management unit and an association maintenance unit; The identification management unit generates a unique identifier for the object model and standardizes the device identifiers in the collected data; the association maintenance unit calculates the correlation between the device identifier and the object model identifier, establishes and maintains the mapping relationship between the two, and handles many-to-many mapping conflicts and mapping migration during version updates.

9. A smart space IoT management and operation system based on a thing model according to claim 6, characterized in that: The fault diagnosis and analysis module includes a real-time anomaly determination unit and a root cause matching unit. The real-time anomaly determination unit monitors fault alarms in real time, extracts real-time operating data of the equipment and compares it with the rated parameters in the object model, calculates the parameter deviation and determines the anomaly level; the root cause matching unit retrieves matching fault features in the mapping relationship of the object model based on the set of abnormal parameters, and outputs the root cause of the fault and related component information in combination with environmental adaptability.

10. A smart space IoT management and operation system based on a thing model according to claim 6, characterized in that: The parameter optimization and verification module includes a feature calculation unit and a closed-loop verification unit; The feature calculation unit extracts feature vectors from periodic operation and maintenance data, calculates comprehensive feature values, and solves for the optimal equipment operation parameter adjustment values ​​through an iterative optimization algorithm; the closed-loop verification unit sends the adjustment parameters to the equipment, monitors the equipment operation status to verify the adjustment effect, and determines whether to update the object model parameters according to the preset compliance standards.

Images