Internet of Things informatization management system for special equipment

Abstract

The invention relates to the technical field of Internet of Things safety management, in particular to a special equipment Internet of Things informatization management system which comprises an assembly reading module, an inspection evolution module, a difference recognition module, a manufacturing hooking module and a credible screening module. According to the method, the initial value of the equipment identification chain is generated through the dynamic splicing assembly data and the installation record, the traceability of the full-life-cycle data of the equipment is enhanced, the dynamic characteristics of the multi-period connection parameters and the function state are extracted by adopting the long-short-term memory network, the running state jump of the equipment is identified, and the real-time performance and continuity of anomaly detection are improved; constructing a data comparison mechanism, analyzing the difference between a manufacturing timestamp and a functional state, forming equipment degradation path visualization, enhancing maintenance decision data support, fusing multi-dimensional data, breaking through the limitation of single-dimensional analysis, and improving the safety management predictive maintenance capability of special equipment through time sequence feature mining and data collaborative verification; and the systematic risk caused by state mutation is reduced.

Description

Technical Field

[0001] The present invention relates to the technical field of Internet of Things security management, and in particular to an Internet of Things information management system for special equipment. Background Art

[0002] IoT information management for special equipment refers to an information system used to collect status information, monitor operating parameters, implement real-time communication and transmission, and centrally manage data. Addressing the complex operating conditions and stringent management requirements of special equipment during use, this system includes sensor-based collection of key operating parameters such as temperature, pressure, and displacement, local data aggregation using short-range wireless communications, and data upload to a remote management platform via cellular networks. Furthermore, it incorporates device identification mechanisms to implement data ownership management, and utilizes timestamp-based recording to form a complete equipment operation log record chain.

[0003] Existing technologies rely on static equipment identification mechanisms, which are difficult to adapt to the dynamic changes in parameters during the operation of special equipment, resulting in a lag in equipment status assessment. Anomaly monitoring methods based on fixed thresholds cannot effectively identify progressive functional degradation phenomena, including slow degradation processes such as the expansion of microcracks in pressure vessels, which can be easily misjudged as normal fluctuations. Discrete data recording methods sever the temporal correlation of equipment status evolution, resulting in a lack of continuous data support for equipment life cycle management. This is typically manifested in the difficulty of forming an effective mapping between equipment maintenance records and real-time operating data. Network transmission security protocols focus on protecting the communication process, but have not established an integrity verification mechanism for the data generation link on the device side, posing a risk of sensor data being maliciously tampered with. The centralized management platform adopts a unified security policy for heterogeneous equipment, ignoring the impact of differences in the operating environments of multiple devices on security control, including the failure to effectively distinguish between the encryption strength requirements of high-temperature and high-pressure equipment and conventional equipment. Summary of the Invention

[0004] The purpose of the present invention is to solve the shortcomings of the prior art and to propose an Internet of Things information management system for special equipment.

[0005] In order to achieve the above object, the present invention adopts the following technical solutions: A special equipment Internet of Things information management system includes: The assembly reading module reads the special equipment assembly data and installation records, obtains the special equipment identification number field and the function status field, performs field splicing and registers them to the equipment management, generates the initial value of the special equipment identification chain and passes it to the inspection evolution module; The inspection evolution module uses the initial value of the special equipment identification chain as an index to obtain the connection parameters and functional status fields of the special equipment in multiple time periods, extracts dynamic features of the fields through the long short-term memory network, identifies the jump fields and writes them into the equipment inspection, generates the equipment state evolution node set and passes it to the difference recognition module; A difference identification module, based on the equipment state evolution node set, constructs a before-and-after state data pair, obtains the manufacturing timestamp field and the function state field, determines the field difference, writes it into the equipment maintenance management, generates an equipment replacement difference item comparison table, and transmits it to the manufacturing attachment module; The manufacturing attachment module obtains the associated manufacturing number, function status field, room temperature record field, and power consumption field based on the equipment state evolution node set and the equipment replacement difference item comparison table, summarizes them, and splices them to generate the equipment life cycle chain structure; The trusted screening module obtains the operation record field, status feedback field, room temperature record field and power consumption field in the operation and maintenance data based on the equipment life cycle chain structure, performs link continuity judgment through k-fold cross-validation, generates an optimized equipment life cycle chain structure and writes it into maintenance management and energy consumption monitoring.

[0006] As a further solution of the present invention, the initial value of the special equipment identification chain is specifically the equipment identification number, the management registration number, and the equipment number; the equipment state evolution node set is specifically the equipment connection parameters, the time period identification value, and the field jump record; the equipment replacement difference item comparison table is specifically the function status field, the judgment difference field, and the difference comparison data; the equipment life cycle chain structure is specifically the manufacturing number, the power consumption field, and the equipment time series; the optimized equipment life cycle chain structure includes the operation record field, the status feedback field, the room temperature record field, the field node sequence, and the link matching degree.

[0007] As a further solution of the present invention, the assembly reading module includes: The equipment collection submodule obtains the assembly data and installation records of special equipment based on the interface of the equipment management system, extracts the equipment identification number and function status fields, and filters the updated data to obtain preliminary equipment assembly records; A field splicing submodule performs field splicing based on the equipment identification number and function status field in the preliminary equipment assembly record, performs formatting processing, and registers the spliced ​​result into the equipment management system to obtain an equipment identification splicing result; The identification chain generation submodule matches the unique identification of the equipment with the initial identification chain format set by the system according to the equipment identification splicing result, and generates the initial value of the special equipment identification chain.

[0008] As a further solution of the present invention, the inspection evolution module includes: An information extraction submodule retrieves and arranges the connection parameters and function status fields of the special equipment within the associated time period based on the initial value of the special equipment identification chain, calculates the change ratio between the fields in multiple time periods, marks the field change trend, and generates the field change rate; The feature recognition submodule calculates the feature strength value through the long short-term memory network according to the field change rate, extracts the field jump period, determines whether the field change belongs to the jump state and screens it to generate key morphological features; The state node submodule calls the key morphological features, extracts the start and end time and change amplitude value corresponding to each jump field, classifies the continuously changing fields into the same node group, and generates a device state evolution node set.

[0009] As a further solution of the present invention, the feature strength value is calculated by the long short-term memory network , using the formula: ; in, Represents the rate of change of the field at the current time step, Indicates the maximum rate of change within the sliding window period, Representative The time span of the sub-hopping cycle, Indicates the maximum value of the jump cycle time span, Indicates the sliding window period The accumulation operation of time steps, Represents a smoothing constant to avoid a denominator of zero, with a value of 0.1. Represents the total number of time steps in the sliding window.

[0010] As a further solution of the present invention, the difference recognition module includes: A state construction submodule, based on the device state evolution node set and dividing the standard to construct the before and after state data pairs, extracts the manufacturing timestamp field and the functional state field in each data pair and maps them to generate a state sequence map; The difference judgment submodule calls the state sequence mapping, calculates the time value difference of the manufacturing timestamp fields at adjacent time points, extracts the degree of change of the state value of the function state field in the field pair, determines whether there is an attribute mutation in the function field, and generates a field jump identification value; The difference item integration submodule calls the field jump identification quantity, combines and compares the manufacturing timestamp field and the functional status field in the data pair to which the jump field belongs, extracts the field group difference category and difference direction and assigns them to equipment maintenance management, and generates an equipment replacement difference item comparison table.

[0011] As a further solution of the present invention, the manufacturing and mounting module includes: The field analysis submodule obtains the manufacturing number, function status field, room temperature record field, and power consumption field corresponding to each node based on the device state evolution node set and the device replacement difference item comparison table, and generates a field attribution result by combining the time span and device number of the node where the field is located; The horizontal splicing submodule calls the field attribution result, horizontally splices the function status field, room temperature record field, and power consumption field of multiple time periods under the same manufacturing number, and extracts the continuous value sequence and interval time period to generate the field time series splicing amount; The life cycle generation submodule calls the field timing splicing quantity, extracts the state sequence length and field change times corresponding to the continuous field group according to the manufacturing number, calculates the splicing coverage period and the continuous change amplitude of the state field during the evolution process of the field structure and integrates them to generate the equipment life cycle chain structure.

[0012] As a further solution of the present invention, the trusted screening module includes: The field extraction submodule obtains the device lifecycle chain structure, extracts the operation record field, status feedback field, room temperature record field, and power consumption field in the operation and maintenance data, filters the fields, calculates the correspondence measurement between the field coverage period and the time period, and generates the node field matching result; The link judgment submodule calls the node field matching results, reconstructs the link connection sequence for the continuous record segments in the operation record field and the status feedback field, calculates the feature strength value through k-fold cross validation, and detects the field jump rate of multiple link structures to generate the link continuation matching rate; The structure optimization submodule selects the device field group whose link connectivity coefficient is greater than the link stability threshold according to the link continuation matching rate, reorganizes the status field and the continuous segment interval structure, and writes maintenance management and energy consumption monitoring to generate an optimized device life cycle chain structure; The link stability threshold is set according to the distribution characteristics of the link connectivity coefficient, and the 70th percentile value is taken as the threshold.

[0013] As a further solution of the present invention, the link connectivity coefficient is calculated by k-fold cross validation. , using the formula: ; in, Indicates the Links in The average connectivity rate in the fold verification, represents the weighted average of all link connectivity rates, represents the normalized standard deviation of the connectivity distribution, Indicates the Normalized weight coefficient of the fold validation set, Indicates the The link in Stability attenuation factor in fold verification, , Indicates the Discount verification time window length and benchmark duration The difference, Indicates the base time set by the system. Indicates the total number of links to be verified. represents the number of cross validation folds, Indicates the benchmark validation fold.

[0014] Compared with the prior art, the advantages and positive effects of the present invention are: In the present invention, the initial value of the equipment identification chain is generated by dynamically splicing assembly data and installation records, and an equipment status index system based on the time dimension is established to enhance the traceability of the equipment's full life cycle data. A long short-term memory network is used to perform dynamic feature extraction on multi-period connection parameters and functional status, to achieve accurate identification of equipment operating status jump behavior, and effectively improve the real-time and continuity of anomaly detection. A before-and-after data comparison mechanism is constructed based on the state evolution node set, and combined with the manufacturing timestamp and functional status difference analysis, a visual expression of the equipment degradation path is formed to enhance the data support capability of maintenance decisions. Through multi-dimensional data fusion, manufacturing parameters, environmental variables and energy consumption indicators are incorporated into a unified analysis framework, and a model for the association of equipment state evolution with external factors is constructed, breaking through the limitations of traditional single-dimensional data analysis. This processing logic enhances the predictive maintenance capability of special equipment safety management and reduces systemic risks caused by state mutations through time series feature mining and multi-source data collaborative verification. BRIEF DESCRIPTION OF THE DRAWINGS

[0015] Figure 1 It is a system flow chart of the present invention.

[0016] Figure 2 This is a system block diagram of the present invention. DETAILED DESCRIPTION

[0017] In order to make the purpose, technical solutions and advantages of the present invention more clearly understood, the present invention will be further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present invention and are not intended to limit the present invention.

[0018] In the description of the present invention, it should be understood that the terms "length," "width," "up," "down," "front," "back," "left," "right," "vertical," "horizontal," "top," "bottom," "inside," "outside," and the like, indicating positions or relationships, are based on the positions or relationships shown in the accompanying drawings and are intended only to facilitate the description of the present invention and simplify the description. They do not indicate or imply that the devices or elements referred to must have a specific orientation, be constructed, or operate in a specific orientation. Therefore, they should not be construed as limiting the present invention. Furthermore, in the description of the present invention, "plurality" means two or more, unless otherwise expressly and specifically defined.

[0019] See also Figure 1 , a special equipment Internet of Things information management system includes: The assembly reading module reads the special equipment assembly data and installation records, obtains the special equipment identification number field and the function status field, performs field splicing and registers them to the equipment management, generates the initial value of the special equipment identification chain and passes it to the inspection evolution module; The inspection evolution module uses the initial value of the special equipment identification chain as an index to obtain the connection parameters and functional status fields of the special equipment in multiple time periods. It uses the long short-term memory network to extract dynamic features of the fields, identifies the jump fields and writes them into the equipment inspection, generates the equipment state evolution node set and passes it to the difference recognition module; The difference identification module constructs the before and after state data pairs based on the equipment state evolution node set, obtains the manufacturing timestamp field and the function state field, determines the field differences, writes them into the equipment maintenance management, generates the equipment replacement difference item comparison table, and passes it to the manufacturing attachment module; The manufacturing attachment module obtains the associated manufacturing number, function status field, room temperature record field, and power consumption field based on the equipment state evolution node set and the equipment replacement difference item comparison table, summarizes them, and splices them to generate the equipment life cycle chain structure; The trusted screening module, based on the equipment lifecycle chain structure, obtains the operation record field, status feedback field, room temperature record field and power consumption field in the operation and maintenance data, performs link continuity judgment through k-fold cross-validation, generates an optimized equipment lifecycle chain structure and writes it into maintenance management and energy consumption monitoring.

[0020] The initial value of the special equipment identification chain is specifically the equipment identification number, management registration number, and equipment number. The equipment status evolution node set is specifically the equipment connection parameters, time period identification value, and field jump record. The equipment replacement difference item comparison table is specifically the functional status field, judgment difference field, and difference comparison data. The equipment life cycle chain structure is specifically the manufacturing number, power consumption field, and equipment time series. The optimized equipment life cycle chain structure includes the operation record field, status feedback field, room temperature record field, field node sequence, and link matching degree.

[0021] See also Figure 2 , the assembly reading module includes: The equipment collection submodule obtains the assembly data and installation records of special equipment based on the interface of the equipment management system, extracts the equipment identification number and function status fields, and filters the updated data to obtain preliminary equipment assembly records; Through the RESTful API interface provided by the equipment management system, the assembly data of special equipment is called regularly every day. The interface obtains JSON format response data and extracts the device identification number. The parsing includes 、 、 The data collection of the field, set the time filter condition to (24-hour time window), compare when traversing data Field, when Keep records, including single device , , (corresponding to 06-27, 00:00:00 UTC), if the current system time is (06-28, 00:00:00 UTC), the record is retained and the output data table includes the following: Table 1 Equipment status data table Equipment Number Status Code Last updated timestamp CN-EX-0627-001 201 1719475200 CN-EX-0627-002 305 1719446400 As shown in Table 1, the status code mapping rules are: , ,pass (06-27, 08:00:00 UTC) Filter valid data.

[0022] The field splicing submodule performs field splicing based on the equipment identification number and function status field in the preliminary equipment assembly record, formats the spliced ​​data, and registers the spliced ​​data into the equipment management system to obtain the equipment identification splicing result. Will Convert to state description according to the mapping table and generate using string interpolation Format, including , set the length check function to ,when By analyzing the first 40 characters, , perform field spelling, and submit it to the server via HTTPPATCH request. Interface, register the spliced ​​results into the device management system, the request body includes .

[0023] The identification chain generation submodule matches the unique identification of the equipment with the initial identification chain format set by the system according to the equipment identification splicing result, and generates the initial value of the special equipment identification chain; By Template Generate identification chain, extract Chinese Date With serial number ,generate , calculate the serial number check code when checking uniqueness: , including serial number correspond , calculated by ,generate , update to the device metadata field .

[0024] See also Figure 2 , the inspection evolution module includes: The information extraction submodule retrieves the connection parameters and function status fields of the special equipment in the associated time period based on the initial value of the special equipment identification chain and arranges them, calculates the change ratio between the fields in multiple time periods, marks the field change trend, and generates the field change rate; Based on the initial value of the identity chain , retrieve the connection parameters and function status fields of special equipment in the associated period Internal equipment and Field, by time step Divide the data into segments and calculate the change ratio of the field values ​​in adjacent time periods , including pressure parameters in At that time , At that time ,but , set thresholds when marking changing trends ,when When marked as sudden increase or sudden decrease, including those shown in Table 2: Table 2 Time period parameter changes Timestamp Pressure value (MPa) Rate of change Trend Marker 1719482400 2.5 - - 1719486000 2.8 0.12 gentle 1719489600 3.2 0.14 gentle 1719493200 3.8 0.19 sudden increase As shown in Table 2, when hour , trigger trend marking, Represents the rate of change of field values ​​and generates a change rate sequence .

[0025] The feature recognition submodule calculates the feature strength value based on the field change rate through the long short-term memory network, extracts the field jump period, determines whether the field change belongs to the jump state, and screens it to generate key morphological features; Input the change rate sequence into the LSTM network and calculate the feature intensity value according to the formula , Represents the rate of change of the field in the current time step, which is calculated by the difference between the parameters of adjacent time steps. The pressure value is At that time , At that time ,but , Represents the maximum value of the rate of change within the sliding window period, taking all The maximum value of the example window period includes ,but , Representative The time span of a jump cycle is calculated by the difference between the jump start and end timestamps. The example represents the jump start time. , end time , then the calculation formula , Represents the maximum value of the jump cycle time span in the previous data, which is set according to the equipment operation record. , Represents the total number of sliding window time steps, which is set according to the hourly data sampling frequency. Represents the smoothing constant to avoid the denominator being zero, set based on engineering experience, and the numerator represents Normalize and square the current rate of change to eliminate the dimension effect. Avoid the division by zero error and substitute into the formula: ; Intensity Threshold , based on the effective jump events in previous data Value distribution setting (reasonable interval representative ), example results , determined as a valid jump event.

[0026] The state node submodule calls key morphological features to extract the start and end time and change amplitude values ​​corresponding to each jump field, classifies the continuously changing fields into the same node group, and generates a device state evolution node set; Extract the start and end time of the transition event (06-27, 09:00:00) and end time (09:40:00), calculate the change range , when the change direction of consecutive periods is the same and When grouped into the same node group, a node set is generated .

[0027] See also Figure 2, the difference recognition module includes: The state construction submodule constructs the before and after state data pairs based on the equipment state evolution node set and the division standard, extracts the manufacturing timestamp field and the functional state field in each data pair, maps them, and generates a state sequence map; Evolving node sets based on device status , construct data pairs in time order ,extract Fields (including )and Field (including 201), establish a mapping relationship , including when the time difference between adjacent nodes When , the sequence is generated , including those shown in Table 3: Table 3 State sequence mapping table Timestamp Status Code Pressure value (MPa) 1719493200 201 3.2 1719496800 305 3.8 As shown in Table 3, the status code conversion rules are: , , complete the state sequence .

[0028] The difference judgment submodule calls the state sequence mapping, calculates the time value difference of the manufacturing timestamp fields at adjacent time points, extracts the degree of change in the state value of the functional state field in the field pair, determines whether there is an attribute mutation in the functional field, and generates a field jump recognition value; Calculate the difference between adjacent timestamps , set the time difference threshold (Set according to equipment maintenance cycle), when Trigger detection in seconds and extract the change in status value , set the state mutation threshold (Based on previous failure data statistics), , judge whether there is a property mutation in the function field, and generate a jump recognition quantity .

[0029] The difference item integration submodule calls the field jump identification parameter to perform a combined comparison of the manufacturing timestamp field and the functional status field in the data pair to which the jump field belongs. It extracts the difference category and difference direction of the field group and assigns them to equipment maintenance management, generating an equipment replacement difference item comparison table. Comparison of hopping field groups , calculate the difference direction , set the difference category threshold ,when When it is classified as a Class A mutation, the maintenance strategy mapping relationship is extracted , perform combined comparison and generate comparison table entries , mutation time: 1719496800, difference category: Class A, maintenance measures: pressure valve maintenance, mutation time: 1719504000, difference category: Class B, maintenance measures: sensor calibration, Class A difference corresponds to the working condition where the pressure parameter change rate exceeds 15%, triggering the target maintenance process.

[0030] See also Figure 2 , manufacturing attachment modules include: The field analysis submodule, based on the equipment state evolution node set and the equipment replacement difference item comparison table, obtains the manufacturing number, function status field, room temperature record field, and power consumption field corresponding to each node, and combines the time span and equipment number of the node where the field is located to generate the field attribution result; Evolving node sets based on device status and the difference table entries , and extract the manufacturing number associated (including 305), (including 25.3℃), (including 12.8kWh) field, by time span (1 hour) Generate attribution records, including when the room temperature fluctuates within the period When As attributed values, they include those shown in Table 4: Table 4 Field attribution results table Manufacturing number Time span Status Code Room temperature Electricity consumption CN-EX-0627-001 1719493200-1719496800 305 25.15 12.8 CN-EX-0627-002 1719500400-1719504000 201 24.7 9.5 As shown in Table 4, the time span is calculated by the difference between the start and end timestamps, the room temperature takes the average of the monitoring values ​​within the period, and the power consumption is the total power consumption in the period.

[0031] The horizontal splicing submodule calls the field attribution results, horizontally splices the function status fields, room temperature record fields, and power consumption fields of multiple time periods under the same manufacturing number, and extracts the continuous value sequence and interval time period to generate the field time series splicing volume; Call the field attribution result, for the same manufacturing number The data of multiple time periods are sorted by time, and the fields are horizontally spliced ​​to generate a sequence: , , , extract the sequence of consecutive values ​​and calculate the interval between consecutive periods seconds, when the interval exceeds the threshold Insert a null value mark when the next period starts at 1719504000, then the interval between consecutive periods Seconds, trigger null value insertion, generate time series .

[0032] The lifecycle generation submodule calls the field timing splicing quantity, extracts the state sequence length and field change count corresponding to the continuous field group based on the manufacturing number, calculates the splicing coverage period and the continuous change range of the state field during the evolution process, and integrates them to generate the equipment lifecycle chain structure; Statistical manufacturing number The length of the state sequence , number of changes (from 201→305), calculate the coverage period (6 hours), calculate the mean change , integrating and generating life cycle chain structure ,in Rationality verification: According to the equipment maintenance standard GB / T19001, the continuous monitoring cycle should be ≥ 4 hours. Hours meet the requirements.

[0033] See also Figure 2 , the trusted filtering module includes: The operation field extraction submodule obtains the equipment lifecycle chain structure, extracts the operation record field, status feedback field, room temperature record field, and power consumption field in the operation and maintenance data, filters the fields, calculates the correspondence measurement between the field coverage period and time period, and generates the node field matching results; Based on the equipment life cycle chain structure , extract the operation record field op-log = [06-27; 09:00:00 start, 09:30:00 pressure adjustment], status feedback field , room temperature record , power consumption , calculate the time period correspondence measure ,when Keep records, measure and filter the fields that meet the requirements according to the field coverage period and time period correspondence, and generate matching results , including those shown in Table 5: Table 5 Node field matching table Timestamp Operation Record Status Code Room temperature (℃) Power consumption (kWh) Match 1719493200 start up 305 25.3 12.8 1.0 1719496800 Pressure regulation 201 24.8 10.2 1.0 As shown in Table 5, the matching degree calculation rule is: field completeness = number of valid fields / total number of fields (4 items), threshold Set according to data quality standard ISO8000.

[0034] The link judgment submodule calls the node field matching results, reconstructs the link connection sequence for the continuous record segments in the operation record field and the status feedback field, calculates the feature strength value through k-fold cross-validation, and detects the field jump rate of multiple link structures to generate the link continuation matching rate; Perform 5-fold cross validation on the operation record and status feedback link ( ), divide the time window is 5 equal-length intervals, Representative The average connectivity rate of the links in the k-fold verification is calculated based on historical data as the number of valid connections / total number of verifications. For example, link 1 successfully connected 4 times in the 5-fold verification, then , Represents the weighted average of all link connectivity rates, calculated as , weight According to the proportion of time window length, the example represents the weight of 3 links: , connectivity rate ,but , Represents the normalized standard deviation, which is calculated as , the example represents , ,but , Representative The normalized weights of the fold validation set are distributed according to the ratio of the time window length to meet , the time window lengths are seconds, then , Represents the stability attenuation factor, which is calculated based on the number of interruptions. The formula is: , the example represents that link 1 is not interrupted in the first fold verification, then , Represents the verification time window and benchmark duration The difference in seconds, the example represents the first fold time window is seconds, then , the calculation example represents the parameter setting , , , Seconds, substitute into the formula to calculate: , calculate the coefficient component of link 1 at the first fold, , after folding and summing, , set the link stability threshold (Take the 70% percentile of previous data), because , it is determined that link 1 is valid.

[0035] The structure optimization generation submodule selects the device field group whose link connectivity coefficient is greater than the link stability threshold based on the link continuation matching rate, reorganizes the status field and continuous segment interval structure, and writes maintenance management and energy consumption monitoring to generate an optimized device life cycle chain structure; Filtered links, where Represents the link connectivity coefficient, and the reorganization status field is , the continuous segment interval is , the field format written into the maintenance management system is , generating an optimized chain structure .

[0036] The above are merely preferred embodiments of the present invention and do not limit the present invention in any other form. Any technician familiar with the profession may use the technical content disclosed above to change or modify it into an equivalent embodiment with equivalent changes and apply it to other fields. However, any simple modification, equivalent change and modification made to the above embodiment based on the technical essence of the present invention without departing from the content of the technical solution of the present invention shall still fall within the scope of protection of the technical solution of the present invention.

Claims

1. A special equipment Internet of Things information management system, characterized in that: The system comprises: The assembly reading module reads the special equipment assembly data and installation records, obtains the special equipment identification number field and the function status field, performs field splicing and registers them to the equipment management, generates the initial value of the special equipment identification chain and passes it to the inspection evolution module; The inspection evolution module uses the initial value of the special equipment identification chain as an index to obtain the connection parameters and functional status fields of the special equipment in multiple time periods, extracts dynamic features of the fields through the long short-term memory network, identifies the jump fields and writes them into the equipment inspection, generates the equipment state evolution node set and passes it to the difference recognition module; A difference identification module, based on the equipment state evolution node set, constructs a before-and-after state data pair, obtains the manufacturing timestamp field and the function state field, determines the field difference, writes it into the equipment maintenance management, generates an equipment replacement difference item comparison table, and transmits it to the manufacturing attachment module; The manufacturing attachment module obtains the associated manufacturing number, function status field, room temperature record field and power consumption field based on the equipment state evolution node set and the equipment replacement difference item comparison table, summarizes them, and splices them to generate the equipment life cycle chain structure.

2. The special equipment Internet of Things information management system according to claim 1, characterized in that: The initial value of the special equipment identification chain is specifically the equipment identification number, management registration number, and equipment number; the equipment state evolution node set is specifically the equipment connection parameters, time period identification value, and field jump record; the equipment replacement difference item comparison table is specifically the function status field, judgment difference field, and difference comparison data; the equipment life cycle chain structure is specifically the manufacturing number, power consumption field, and equipment time series.

3. The special equipment Internet of Things information management system according to claim 1, characterized in that: The assembly reading module comprises: The equipment collection submodule obtains the assembly data and installation records of special equipment based on the interface of the equipment management system, extracts the equipment identification number and function status fields, and filters the updated data to obtain preliminary equipment assembly records; A field splicing submodule performs field splicing based on the equipment identification number and function status field in the preliminary equipment assembly record, performs formatting processing, and registers the spliced ​​result into the equipment management system to obtain an equipment identification splicing result; The identification chain generation submodule matches the unique identification of the equipment with the initial identification chain format set by the system according to the equipment identification splicing result, and generates the initial value of the special equipment identification chain.

4. The special equipment Internet of Things information management system according to claim 1, characterized in that: The inspection evolution module includes: An information extraction submodule retrieves and arranges the connection parameters and function status fields of the special equipment within the associated time period based on the initial value of the special equipment identification chain, calculates the change ratio between the fields in multiple time periods, marks the field change trend, and generates the field change rate; The feature recognition submodule calculates the feature strength value through the long short-term memory network according to the field change rate, extracts the field jump period, determines whether the field change belongs to the jump state and screens it to generate key morphological features; The state node submodule calls the key morphological features, extracts the start and end time and change amplitude value corresponding to each jump field, classifies the continuously changing fields into the same node group, and generates a device state evolution node set.

5. The special equipment Internet of Things information management system according to claim 4, characterized in that: The feature strength value is calculated by the long short-term memory network , using the formula: ； in, Represents the rate of change of the field at the current time step, Indicates the maximum rate of change within the sliding window period, Representative The time span of the sub-hopping cycle, Indicates the maximum value of the jump cycle time span, Indicates the sliding window period The accumulation operation of time steps, Represents a smoothing constant to avoid a denominator of zero, with a value of 0.

1. Represents the total number of time steps in the sliding window.

6. The special equipment Internet of Things information management system according to claim 1, characterized in that: The difference recognition module includes: A state construction submodule, based on the device state evolution node set and dividing the standard to construct the before and after state data pairs, extracts the manufacturing timestamp field and the functional state field in each data pair and maps them to generate a state sequence map; The difference judgment submodule calls the state sequence mapping, calculates the time value difference of the manufacturing timestamp fields at adjacent time points, extracts the degree of change of the state value of the function state field in the field pair, determines whether there is an attribute mutation in the function field, and generates a field jump identification value; The difference item integration submodule calls the field jump identification quantity, combines and compares the manufacturing timestamp field and the functional status field in the data pair to which the jump field belongs, extracts the field group difference category and difference direction and assigns them to equipment maintenance management, and generates an equipment replacement difference item comparison table.

7. The special equipment Internet of Things information management system according to claim 1, characterized in that: The manufacturing attachment module includes: The field analysis submodule obtains the manufacturing number, function status field, room temperature record field, and power consumption field corresponding to each node based on the device state evolution node set and the device replacement difference item comparison table, and generates a field attribution result by combining the time span and device number of the node where the field is located; The horizontal splicing submodule calls the field attribution result, horizontally splices the function status field, room temperature record field, and power consumption field of multiple time periods under the same manufacturing number, and extracts the continuous value sequence and interval time period to generate the field time series splicing amount; The life cycle generation submodule calls the field timing splicing quantity, extracts the state sequence length and field change times corresponding to the continuous field group according to the manufacturing number, calculates the splicing coverage period and the continuous change amplitude of the state field during the evolution process of the field structure and integrates them to generate the equipment life cycle chain structure.

8. The special equipment Internet of Things information management system according to claim 1, characterized in that: The system further comprises: The trusted screening module, based on the device lifecycle chain structure, obtains the operation record field, status feedback field, room temperature record field, and power consumption field in the operation and maintenance data, performs link continuity judgment through k-fold cross-validation, generates an optimized device lifecycle chain structure, and writes it into maintenance management and energy consumption monitoring; The optimized device lifecycle chain structure includes an operation record field, a status feedback field, a room temperature record field, a field node sequence, and a link matching degree.

9. The special equipment Internet of Things information management system according to claim 8, characterized in that: The trusted screening module includes: The field extraction submodule obtains the device lifecycle chain structure, extracts the operation record field, status feedback field, room temperature record field, and power consumption field in the operation and maintenance data, filters the fields, calculates the correspondence measurement between the field coverage period and the time period, and generates the node field matching result; The link judgment submodule calls the node field matching results, reconstructs the link connection sequence for the continuous record segments in the operation record field and the status feedback field, calculates the feature strength value through k-fold cross validation, and detects the field jump rate of multiple link structures to generate the link continuation matching rate; The structure optimization submodule selects the device field group whose link connectivity coefficient is greater than the link stability threshold according to the link continuation matching rate, reorganizes the status field and the continuous segment interval structure, and writes maintenance management and energy consumption monitoring to generate an optimized device life cycle chain structure; The link stability threshold is set according to the distribution characteristics of the link connectivity coefficient, and the 70th percentile value is taken as the threshold.

10. The special equipment Internet of Things information management system according to claim 9, characterized in that: The link connectivity coefficient is calculated by k-fold cross validation , using the formula: ； in, Indicates the Links in The average connectivity rate in the fold verification, represents the weighted average of all link connectivity rates, represents the normalized standard deviation of the connectivity distribution, Indicates the Normalized weight coefficient of the fold validation set, Indicates the The link in Stability attenuation factor in fold verification, , Indicates the Discount verification time window length and benchmark duration The difference, Indicates the base time set by the system. Indicates the total number of links to be verified. represents the number of cross validation folds, Indicates the benchmark validation fold.

Images