A cloud-based facility management method and system

Abstract

The present invention discloses a cloud-based facility management method and system, which relates to the field of facility management technology, including multiple local modules, multiple edge computing modules and cloud modules. The cloud module establishes a global comparison model based on the environment and production tasks of each facility, and splits the comparison module into multiple sub-models based on the facilities matched by each edge computing module, and the cloud module transmits each sub-model to the edge computing module matched by the corresponding facility; each local module includes multiple monitoring units, each monitoring unit matches a group of facilities respectively, and each monitoring unit obtains the monitoring data of the matching facilities. The present invention reduces data transmission delay and improves abnormality detection and response speed by combining cloud computing and edge computing, and can more accurately identify abnormalities and reduce false positives and false negatives through the influence of dynamic comparison models and the health status of facilities.

Description

Technical Field

[0001] The present invention relates to the technical field of facility management, and in particular to a cloud-based facility management method and system. Background Art

[0002] With the rapid development of industrial automation, Internet of Things (IoT) and intelligent facility management, real-time monitoring and fault detection of facility equipment have become particularly important. As a mature technology, cloud computing is widely used in data storage, processing and analysis;

[0003] The cloud usually acts as a data storage and computing platform, using powerful computing power to perform data analysis, generate diagnostic results, and support remote monitoring and management. The cloud platform can centrally manage monitoring data from different facilities, support remote monitoring, and greatly improve management efficiency. Cloud computing has powerful computing and storage capabilities and can process data from a large number of devices and sensors.

[0004] After searching, a Chinese patent (publication number: CN114553872B) discloses a cloud-based data center infrastructure management system, which includes: a cloud server, which is connected to a sub-server, and is used to establish a connection between sub-servers and determine whether to alarm; a sub-server, which is connected to an infrastructure node, and is used to aggregate sub-server information; and a monitoring device, which is connected to the sub-server.

[0005] In the prior art, facilities usually have a relationship of mutual dependence and collaboration. Traditional systems do not fully consider the dependence and mutual influence between facilities, resulting in inaccurate anomaly detection and source positioning. The fault propagation path between facilities is also not fully captured, which increases the difficulty of fault diagnosis. Therefore, the present invention proposes a cloud-based facility management method and system. Summary of the invention

[0006] The purpose of the present invention is to provide a cloud-based facility management method and system to solve the problems mentioned in the above background technology.

[0007] The present invention can be implemented by the following technical solutions: a cloud-based facility management system, including a plurality of local modules, a plurality of edge computing modules and a cloud module;

[0008] The cloud module establishes a global comparison model based on the environment and production tasks of each facility, and splits the comparison module into multiple sub-models based on the facilities matched by each edge computing module. Each sub-model is customized for the facilities matched by each edge computing module and its production tasks, covering the standard sensor data and fault pre-processing solutions related to the corresponding facilities, and the cloud module transmits each sub-model to the edge computing module matched by the corresponding facility.

[0009] Each local module includes multiple monitoring units, each monitoring unit matches a group of facilities respectively, and each monitoring unit includes multiple sensors, each sensor monitors each data of the matching facilities respectively, and obtains corresponding monitoring data. The monitoring items are selected based on the fluctuation information that may be generated when the facilities are working, including temperature, humidity, pressure, vibration, current, voltage, gas concentration, position, motion state, air quality or flow, etc.;

[0010] After obtaining the monitoring data, each monitoring unit transmits the monitoring data to the corresponding edge computing module;

[0011] Each edge computing module includes a data processing unit and a comparison unit;

[0012] The data processing unit is used to receive the monitoring data transmitted by the corresponding regional monitoring unit and perform preprocessing on it, the preprocessing including data filtering, noise removal, data compression and data encryption of the monitoring data;

[0013] After preprocessing the monitoring data, the data processing unit transmits the monitoring data to the comparison unit and the cloud module;

[0014] The comparison unit is used to receive the sub-model of the corresponding facility transmitted by the cloud module;

[0015] After receiving the monitoring data transmitted by the edge computing module, the cloud module compares it with the comparison model, and then monitors the working status of each facility;

[0016] And after comparing the monitoring data, the cloud module sends an identification signal to the corresponding edge computing module;

[0017] After the edge computing module receives the identification signal, the comparison unit inside it does not work;

[0018] When the edge computing module fails to receive an identification signal within a preset time threshold, its internal comparison unit starts to continue monitoring the monitoring data preprocessed by its internal data processing unit, and when the monitoring data does not match the sub-module, it performs obstacle processing based on the fault preprocessing solution.

[0019] A further technical improvement of the present invention is that the comparison model is provided with a dynamic influence mechanism, which continuously adjusts the comparison model based on the influence factors of the corresponding facilities, the influence factors including the service life, maintenance status and working environment of the facilities, and the specific steps include:

[0020] Step 1: Establish a basic comparison model, which is used to describe the standard working state of the facility under normal working conditions, and takes into account the service life, maintenance records, and environmental impact of the facility, and sets a dynamic adjustment factor Δ(t). The dynamic adjustment factor is used to adjust the standard value of the comparison model according to changes in the service life, working environment, and maintenance status of the facility;

[0021] The comparison model is represented as a multidimensional standard space, denoted as M dw , which contains various standard data under normal operation of the facility;

[0022] The value of the standard data at time t is X dw (t), which is positioned as X dw (t)=[X1(t),X2(t),...X n (t)];

[0023] Where X i (t) represents the standard value of the i-th monitoring data at time t;

[0024] And, Δ(t)=[Δ1(t), Δ2(t),...Δ n (t)],Δ n (t) is the adjustment factor associated with the i-th monitoring data, which is adjusted based on the aging of the facility, environmental changes, or maintenance records;

[0025] By base model X dw (t) and dynamic adjustment factor Δ(t), dynamic comparison model M dt (t) is expressed as: M dt (t) = X dw (t)+Δt(t);

[0026] Step 2: Set the allowable deviation range. To ensure that the working status of the facility is always within a reasonable range, set the allowable deviation range for each monitoring item X. i (t) Setting the allowable deviation range;

[0027] Step 3: Compare the dynamic impact mechanism with the comparison model;

[0028] Transform the facility's real-time monitoring data into jc (t) Compared with the dynamic comparison model, the goal of the dynamic impact mechanism is to ensure real-time monitoring of data X jc (t) Always stay within the tolerance range of the dynamic contrast model;

[0029] The difference of the whole system is measured by setting the vector range: ||X jc (t)-M dt (t)||≤α;

[0030] Among them, ||·|| represents the vector range, and α is the maximum allowed deviation range;

[0031] Step 4: Establish a dynamic adjustment mechanism. jc When the deviation between (t) and the dynamic comparison model exceeds the allowable range α, the cloud module will adjust the dynamic adjustment factor Δ(t) to correct the deviation. The expression formula is:

[0032] Among them, Δ i (t) represents the dynamic adjustment factor of the i-th monitoring data, the adjustment amount at time t, which is used to adjust the parameters of the comparison model or facility according to the difference between the current state of the facility and the standard state

[0033] β is the adjustment coefficient, which indicates the degree of correction and determines the intensity of the adjustment. You can choose a suitable value according to the actual situation, 0≤β≤1. The larger the value, the greater the impact of the adjustment, and the smaller the value, the smaller the impact of the adjustment.

[0034] The actual monitoring data of the ith item at time t, that is, the real-time data obtained from the sensor;

[0035] X i (t) is the standard value of the i-th item at time t, which comes from the comparison model;

[0036] Indicates the deviation between the actual monitoring data and the standard value;

[0037] If the deviation is positive, it means that the actual data is higher than the standard value; if it is negative, it means that the actual data is lower than the standard value.

[0038] A further technical improvement of the present invention is that the cloud module sets the influence range R i (t), represents the influence of the influencing factor on the i-th monitoring data, which is expressed as:

[0039] If R i (t) If the deviation exceeds the allowable range α, the preset early warning mechanism will be triggered; otherwise, it will be regarded as a normal fluctuation;

[0040] Each time the monitoring data is updated, the cloud module monitors the comparison model in real time through the dynamic impact mechanism and calculates the impact range R i (t) and determine whether it exceeds the allowable range.

[0041] A further technical improvement of the present invention is that: the data processing unit performs standardization processing on the monitoring data to ensure that each data is compared on the same scale;

[0042] The formula used is:

[0043] Among them, X i (t) bz is the standard value of the standardized monitoring data of the facility ith at time t;

[0044] X i (t) is the standard value of the monitoring data of the facility ith at time t;

[0045] μ i is the average value of monitoring data of facility i;

[0046] σ i is the standard deviation of the monitoring data of the facility ith item;

[0047] After receiving the monitoring data, the cloud module first determines whether the monitoring data is abnormal based on the comparison model;

[0048] After the cloud module determines that the monitoring data is abnormal, the source of the abnormality is located. The cloud module traces the source of the abnormality based on the relationship between the facilities;

[0049] By using graph theory to model the relationship between various facilities, a facility dependency graph is established. When the monitoring data of a facility is judged to be abnormal, the source is found through the propagation path of the facility dependency graph.

[0050] A further technical improvement of the present invention is that: each node in the facility dependency graph represents a facility, and each edge represents the interdependence between facilities;

[0051] The facility life and maintenance records are recorded as node attributes in the facility dependency graph.

[0052] A further technical improvement of the present invention is that after adding the facility life, based on the current service life T of the corresponding facility d and the maximum lifetime T max Calculate the life weight W of the corresponding facility life ;

[0053]

[0054] After adding the maintenance records, according to the total number of maintenance times R of the corresponding facility d and the expected maximum number of maintenance times R max , calculate the maintenance weight W r ;

[0055]

[0056] By formula The age and maintenance records of the facility have an impact on the location of the anomaly source;

[0057] Among them, ω i (t) is the anomaly measure weighted by facility age and maintenance records;

[0058] (1+W life,i +W r,i ) is a weighted adjustment to the anomaly metric, reflecting the life weight and maintenance weight of the facility, which quantifies the health of the facility by reflecting the aging of the equipment and the maintenance history;

[0059] W life,i is the life weight of facility i, which is related to the service life and expected life of the equipment;

[0060] W r,i is the maintenance record weight of facility i, indicating the maintenance frequency and historical maintenance status of the equipment;

[0061] A further technical improvement of the present invention is that when the monitoring data of a facility is determined to be abnormal, the device dependency graph is used to find other facilities related to the facility through its edges and nodes;

[0062] The influence of the lifespan and maintenance records of each facility is taken into account. By dynamically adjusting the equipment anomaly threshold, it is easier to trigger the location of the source of anomalies for equipment with greater impact, such as aging equipment and equipment that is frequently repaired.

[0063] The present invention also discloses a cloud-based facility management method, which comprises the following steps:

[0064] S1. The cloud module establishes a global comparison model, and splits the global comparison model into multiple sub-models based on the environment and production tasks of each facility, and transmits each sub-model to the comparison unit in the corresponding edge computing module;

[0065] S2. The local module equips each facility with a monitoring unit to monitor the working status of each facility, collect monitoring data, and after collecting the corresponding monitoring data, transmit it to the corresponding edge computing module for preprocessing;

[0066] S3, the edge computing module transmits the pre-processed monitoring data to the cloud module and receives the identification signal from the module;

[0067] And the edge computing module stops or starts the operation of its internal comparison unit based on whether the identification signal is received;

[0068] After receiving the recognition signal, the comparison unit inside the edge computing module stops working;

[0069] If no identification signal is received within the preset time threshold, the comparison unit of the edge computing module will be activated to process the monitoring data through its internal sub-model;

[0070] S4. The cloud module receives the monitoring data from the edge computing module and compares it with the pre-defined comparison model to detect whether the facility is working properly.

[0071] Compared with the prior art, the present invention has the following beneficial effects:

[0072] The present invention combines cloud computing and edge computing to reduce data transmission delays, improve anomaly detection and response speed, and can more accurately identify anomalies and reduce false positives and missed positives by dynamically comparing the impact of models and facility health status;

[0073] At the same time, by introducing the device dependency graph, it can help to quickly trace the source of the abnormality, shorten the fault diagnosis and processing time, and by dynamically adjusting the threshold, the system can automatically adapt to factors such as equipment aging and environmental changes, thereby improving the level of intelligent equipment management. BRIEF DESCRIPTION OF THE DRAWINGS

[0074] In order to facilitate understanding by those skilled in the art, the present invention is further described below with reference to the accompanying drawings.

[0075] Figure 1 It is a flow chart of the cloud-based facility management method in the present invention. DETAILED DESCRIPTION

[0076] In order to further explain the technical means and effects adopted by the present invention to achieve the predetermined invention purpose, the specific implementation methods, structures, features and effects of the present invention are described in detail below in conjunction with the accompanying drawings and preferred embodiments.

[0077] See also Figure 1 As shown, the present invention provides a cloud-based facility management system, including multiple local modules, multiple edge computing modules and a cloud module;

[0078] The cloud module establishes a global comparison model based on the environment and production tasks of each facility, and splits the comparison module into multiple sub-models based on the facilities matched by each edge computing module. Each sub-model is customized for the facilities and production tasks matched by each edge computing module, covering the standard sensor data and fault pre-processing solutions related to the corresponding facilities, and the cloud module transmits each sub-model to the edge computing module matched by the corresponding facility.

[0079] The comparison model is equipped with a dynamic impact mechanism, which continues to adjust the comparison model based on the influencing factors of the corresponding facilities. The influencing factors include the service life of the facilities, maintenance status, and working environment. The specific steps include:

[0080] Step 1: Establish a basic comparison model, which is used to describe the standard working state of the facility under normal working conditions, and takes into account the service life, maintenance records, and environmental impact of the facility, and sets a dynamic adjustment factor Δ(t). The dynamic adjustment factor is used to adjust the standard value of the comparison model according to changes in the service life, working environment, and maintenance status of the facility;

[0081] The comparison model is represented as a multidimensional standard space, denoted as M dw , which contains various standard data under normal operation of the facility, such as temperature, humidity, pressure, vibration, current, voltage, gas concentration, position, movement status, air quality, and flow;

[0082] The value of the standard data at time t is X dw (t), which is located at X dw (t)=[X1(t),X2(t),...X n (t)];

[0083] Where X i (t) represents the standard value of the i-th monitoring data at time t;

[0084] And, Δ(t)=[Δ1(t), Δ2(t),...Δ n (t)],Δ n (t) is the adjustment factor associated with the i-th monitoring data, which is adjusted based on the aging of the facility, environmental changes, or maintenance records;

[0085] By base model X dw (t) and dynamic adjustment factor Δ(t), dynamic comparison model M dt (t) is expressed as: M dt (t) = X dw (t)+Δt(t);

[0086] Step 2: Set the allowable deviation range. To ensure that the working status of the facility is always within a reasonable range, set the allowable deviation range for each monitoring item X. i (t) Setting the allowable deviation range;

[0087] In this embodiment, for the i-th data, the maximum allowed deviation is δ i , then the allowable range of the i-th data is X i (t)∈[X i (t)-δ i ,Xi (t)+δ i ];

[0088] Step 3: Compare the dynamic impact mechanism with the comparison model;

[0089] Transform the facility's real-time monitoring data into jc (t) Compared with the dynamic comparison model, the goal of the dynamic impact mechanism is to ensure real-time monitoring of data X jc (t) Always stay within the tolerance range of the dynamic contrast model;

[0090] The difference of the whole system is measured by setting the vector range: ||X jc (t)-M dt (t)||≤α;

[0091] Among them, ||·|| represents the vector range, and α is the maximum allowed deviation range;

[0092] Step 4: Establish a dynamic adjustment mechanism. jc When the deviation between (t) and the dynamic comparison model exceeds the allowable range α, the cloud module will adjust the dynamic adjustment factor Δ(t) to correct the deviation. The expression formula is:

[0093] Among them, Δ i (t) represents the dynamic adjustment factor of the i-th monitoring data, the adjustment amount at time t, which is used to adjust the parameters of the comparison model or facility according to the difference between the current state of the facility and the standard state

[0094] β is the adjustment coefficient, which indicates the degree of correction and determines the intensity of the adjustment. You can choose a suitable value according to the actual situation, 0≤β≤1. The larger the value, the greater the impact of the adjustment, and the smaller the value, the smaller the impact of the adjustment.

[0095] The actual monitoring data of the ith item at time t, that is, the real-time data obtained from the sensor;

[0096] X i (t) is the standard value of the i-th item at time t, which comes from the comparison model;

[0097] Indicates the deviation between the actual monitoring data and the standard value. If the deviation is positive, it means that the actual data is higher than the standard value; if it is negative, it means that the actual data is lower than the standard value;

[0098] Cloud module sets the impact range R i (t), represents the influence of the influencing factor on the i-th monitoring data, which is expressed as:

[0099] If R i (t) If the deviation exceeds the allowable range α, the preset early warning mechanism will be triggered; otherwise, it will be regarded as a normal fluctuation;

[0100] Each time the monitoring data is updated, the cloud module monitors the comparison model in real time through the dynamic impact mechanism and calculates the impact range R i (t) and determine whether it exceeds the allowable range.

[0101] Each local module includes multiple monitoring units, each monitoring unit matches a group of facilities respectively, and each monitoring unit includes multiple sensors, each sensor monitors each data of the matching facilities respectively, and obtains corresponding monitoring data. The monitoring items are selected based on the fluctuation information that may be generated when the facilities are working, including temperature, humidity, pressure, vibration, current, voltage, gas concentration, position, motion state, air quality, and flow;

[0102] After obtaining the monitoring data, each monitoring unit transmits the monitoring data to the corresponding edge computing module;

[0103] Each edge computing module includes a data processing unit and a comparison unit;

[0104] The data processing unit is used to receive the monitoring data transmitted by the corresponding regional monitoring unit and perform preprocessing on it, the preprocessing including data filtering, noise removal, data compression and data encryption of the monitoring data;

[0105] In this embodiment, data filtering and noise removal use a filtering algorithm to smooth the monitoring data, thereby removing useless data generated by signal fluctuations or external interference and retaining valid information;

[0106] Data compression uses compression algorithms to further compress monitoring data, reducing data volume without losing key information, thereby alleviating bandwidth pressure;

[0107] Data encryption means that before transmission, the data processing unit performs AES encryption on the monitoring data to ensure that the data will not be stolen or tampered with during transmission;

[0108] After preprocessing the monitoring data, the data processing unit transmits the monitoring data to the comparison unit and the cloud module;

[0109] The comparison unit is used to receive the sub-model of the corresponding facility transmitted by the cloud module;

[0110] The data processing unit standardizes the monitoring data to ensure that all data can be compared on the same scale;

[0111] The formula used is:

[0112] Among them, X i (t) bz is the standard value of the standardized monitoring data of the facility ith at time t;

[0113] X i (t) is the standard value of the monitoring data of the facility ith at time t;

[0114] μ i is the average value of monitoring data of facility i;

[0115] σ i is the standard deviation of the monitoring data of the facility ith item;

[0116] After receiving the monitoring data, the cloud module first determines whether the monitoring data is abnormal based on the comparison model. In this embodiment, the mean and standard deviation of the monitoring data are calculated and a threshold is set to identify abnormal data points. That is, when the monitoring data exceeds the multiple of the corresponding standard deviation in the comparison model, the data is considered abnormal.

[0117] After the cloud module determines that the monitoring data is abnormal, the source of the abnormality is located. The cloud module traces the source of the abnormality based on the relationship between the facilities;

[0118] By using graph theory to model the relationship between various facilities, a facility dependency graph is established. When the monitoring data of a facility is judged to be abnormal, the source is found through the propagation path of the facility dependency graph.

[0119] Each node in the facility dependency graph represents a facility, and each edge represents the interdependency between facilities;

[0120] The facility life and maintenance records are recorded as node attributes in the facility dependency graph;

[0121] After adding the facility life, based on the current service life of the corresponding facility, T d and the maximum lifetime T max Calculate the life weight W of the corresponding facility life ;

[0122]

[0123] After adding the maintenance records, according to the total number of maintenance times R of the corresponding facility d and the expected maximum number of maintenance times R max , calculate the maintenance weight W r ;

[0124]

[0125] By formula The age and maintenance records of the facility have an impact on the location of the anomaly source;

[0126] Among them, ω i (t) is the anomaly measure weighted by facility age and maintenance records;

[0127] (1+W life,i +W r,i ) is a weighted adjustment to the anomaly metric, reflecting the life weight and maintenance weight of the facility, which quantifies the health of the facility by reflecting the aging of the equipment and the maintenance history;

[0128] W life,i is the life weight of facility i, which is related to the service life and expected life of the equipment;

[0129] W r,i is the maintenance record weight of facility i, indicating the maintenance frequency and historical maintenance status of the equipment;

[0130] When the monitoring data of a facility is judged to be abnormal, the device dependency graph is used to find other facilities related to the facility through its edges and nodes;

[0131] The influence of the lifespan and maintenance records of each facility is taken into account. By dynamically adjusting the equipment anomaly threshold, it is easier to trigger the location of the source of anomalies for equipment with greater impact, such as aging equipment and equipment that is frequently repaired.

[0132] After receiving the monitoring data transmitted by the edge computing module, the cloud module compares it with the comparison model and then monitors the working status of each facility;

[0133] And after comparing the monitoring data, the cloud module sends an identification signal to the corresponding edge computing module;

[0134] After the edge computing module receives the identification signal, the comparison unit inside it does not work;

[0135] When the edge computing module fails to receive an identification signal within a preset time threshold, its internal comparison unit starts to continue monitoring the monitoring data preprocessed by its internal data processing unit, and when the monitoring data does not match the sub-module, it performs obstacle processing based on the fault preprocessing solution.

[0136] The above description is only a preferred embodiment of the present invention and does not limit the present invention in any form. Although the present invention has been disclosed as a preferred embodiment as above, it is not used to limit the present invention. Any technical personnel in this field can make some changes or modify the technical contents disclosed above into equivalent embodiments without departing from the scope of the technical solution of the present invention. However, any simple modification, equivalent change and modification made to the above embodiments according to the technical essence of the present invention without departing from the content of the technical solution of the present invention still fall within the scope of the technical solution of the present invention.

Claims

1. A cloud-based facility management system, comprising a plurality of local modules, a plurality of edge computing modules and a cloud module, characterized in that: The cloud module establishes a global comparison model based on the environment and production tasks of each facility, and splits the comparison module into multiple sub-models based on the facilities matched by each edge computing module, and the cloud module transmits each sub-model to the edge computing module matched by the corresponding facility; Each local module includes a plurality of monitoring units, each monitoring unit is matched with a group of facilities, and each monitoring unit obtains monitoring data of the matched facilities; After obtaining the monitoring data, each monitoring unit transmits the monitoring data to the corresponding edge computing module; Each edge computing module includes a data processing unit and a comparison unit; The data processing unit is used to receive the monitoring data transmitted by the corresponding area monitoring unit and pre-process it; After preprocessing the monitoring data, the data processing unit transmits the monitoring data to the comparison unit and the cloud module; The comparison unit is used to receive the sub-model of the corresponding facility transmitted by the cloud module; After receiving the monitoring data transmitted by the edge computing module, the cloud module compares it with the comparison model; And after comparing the monitoring data, the cloud module sends an identification signal to the corresponding edge computing module; After the edge computing module receives the identification signal, the comparison unit inside it does not work; When the edge computing module fails to receive an identification signal within a preset time threshold, its internal comparison unit starts to continue monitoring the monitoring data preprocessed by its internal data processing unit, and when the monitoring data does not match the sub-module, it performs obstacle processing based on the fault preprocessing solution.

2. A cloud-based facility management system according to claim 1, characterized in that: The preprocessing of the data processing unit includes data filtering, noise removal, data compression and data encryption of the monitoring data.

3. A cloud-based facility management system according to claim 1, characterized in that: The comparison model is set up with a dynamic impact mechanism, which continues to adjust the comparison model based on the influencing factors of the corresponding facilities; Influencing factors include facility life, maintenance status and working environment.

4. A cloud-based facility management system according to claim 3, characterized in that: The steps of the dynamic impact mechanism include: Step 1: Establish a basic comparison model and set a dynamic adjustment factor. The dynamic adjustment factor is used to adjust the standard value of the comparison model according to changes in the service life, working environment, and maintenance status of the facility; Step 2: Set the allowable deviation range; Step 3: Compare the dynamic impact mechanism with the comparison model; Step 4: Establish a dynamic adjustment mechanism. When the deviation between the actual monitoring data of the facility and the dynamic comparison model exceeds the allowable range, the cloud module will adjust the dynamic adjustment factor to correct the deviation.

5. A cloud-based facility management system according to claim 4, characterized in that: The cloud module sets the impact range, which indicates the degree of influence of the influencing factor on the i-th monitoring data; If the impact range exceeds the allowable deviation range, the preset early warning mechanism will be triggered, otherwise, it will be regarded as a normal fluctuation; Each time the monitoring data is updated, the cloud module monitors the comparison model in real time through a dynamic impact mechanism, calculates the impact range and determines whether it exceeds the allowable range.

6. A cloud-based facility management system according to claim 1, characterized in that: The data processing unit performs standardized processing on the monitoring data; After receiving the monitoring data, the cloud module first determines whether the monitoring data is abnormal based on the comparison model; After the cloud module determines that the monitoring data is abnormal, the source of the abnormality is located. The cloud module traces the source of the abnormality based on the relationship between the facilities; By using graph theory to model the relationship between various facilities, a facility dependency graph is established. When the monitoring data of a facility is judged to be abnormal, the source is found through the propagation path of the facility dependency graph.

7. A cloud-based facility management system according to claim 6, characterized in that: Each node in the facility dependency graph represents a facility, and each edge represents the interdependency between facilities; The facility life and maintenance records are recorded as node attributes in the facility dependency graph.

8. A cloud-based facility management method according to any one of claims 1 to 7, characterized in that: The method comprises the following steps: S1. The cloud module establishes a global comparison model, and splits the global comparison model into multiple sub-models based on the environment and production tasks of each facility, and transmits each sub-model to the comparison unit in the corresponding edge computing module; S2. The local module equips each facility with a monitoring unit to monitor the working status of each facility, collect monitoring data, and after collecting the corresponding monitoring data, transmit it to the corresponding edge computing module for preprocessing; S3, the edge computing module transmits the pre-processed monitoring data to the cloud module and receives the identification signal from the cloud module; And the edge computing module stops or starts the operation of its internal comparison unit based on whether the identification signal is received; After receiving the recognition signal, the comparison unit inside the edge computing module stops working; If no identification signal is received within the preset time threshold, the comparison unit of the edge computing module will be activated to process the monitoring data through its internal sub-model; S4. The cloud module receives the monitoring data from the edge computing module and compares it with the pre-defined comparison model to detect whether the facility is working properly.

Images