Electromechanical equipment information management and control method based on BIM technology

Abstract

The invention relates to the technical field of information management and control, and discloses an electromechanical equipment information management and control method based on a BIM technology, and the method comprises the following steps: S1, deploying a multi-modal energy collection module to form a self-powered sensor network, achieving the continuous collection of equipment data, and providing an original data source; s2, a dynamic protection system is constructed according to the data collected in the S1 to provide a safe data environment, safe encryption and transmission of full-life-cycle information of the equipment are achieved, and the dynamic protection system integrates a BIM model and a chaotic encryption mechanism. According to the invention, the sampling mode of the sensor is automatically switched according to the voltage of the super capacitor, the device data can be effectively acquired in different energy states, complex wiring is not needed, the influence of power supply interruption is avoided, the continuous acquisition of the device data is realized, and the stability and reliability of data acquisition are improved.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of information management and control, in particular to a mechanical and electrical equipment information management and control method based on BIM technology. BACKGROUND

[0002] Mechanical and electrical equipment information management and control refers to a series of management activities that comprehensively use various technical means to collect, transmit, store, analyze and apply various types of information generated in the whole life cycle process of mechanical and electrical equipment from planning and design, production and manufacturing, installation and debugging, operation and maintenance to scrap disposal, in order to realize real-time monitoring of equipment state, fault early warning, performance optimization and scientific decision-making. In the fields of modern industrial production and intelligent building, there are various types of mechanical and electrical equipment with complex structure and variable operating environment. Mechanical and electrical equipment information management and control can help enterprises and management personnel to timely grasp the equipment operation status, ensure stable operation of equipment, improve production efficiency and reduce operating cost, which is of great significance to improve the reliability and safety of the whole system.

[0003] After searching, the following problems still exist in the prior art: 1. Equipment data collection usually relies on externally powered sensors, which not only increases wiring cost and maintenance difficulty, but also is easily affected by power supply interruption in complex environment, resulting in discontinuous data collection and inability to obtain real-time and comprehensive equipment operation status information; 2. The transmission and storage of equipment whole life cycle information lack effective security protection mechanism, and conventional encryption method is difficult to resist complex and variable network attacks, especially when equipment fails, the risk of information leakage increases significantly, which cannot guarantee data security; 3. The traditional model has insufficient feature extraction capability for equipment operation state, and has great delay in capturing transient faults, which makes it difficult to quickly and accurately identify fault types and take effective maintenance measures in time, which is easy to cause equipment damage and production interruption. SUMMARY

[0004] (I) Technical problems solved In view of the deficiencies in the prior art, the present application provides a mechanical and electrical equipment information management and control method based on BIM technology, which solves the problem of "low efficiency" in the above background technology.

[0005] (II) Technical scheme In order to achieve the above purpose, the present application is implemented by the following technical scheme: a mechanical and electrical equipment information management and control method based on BIM technology, comprising the following steps: S1: Deploy a multi-modal energy collection module to form a self-powered sensor network, realize continuous collection of equipment data, and provide original data source; S2: Construct a dynamic protection system according to the data collected in S1 to provide a secure data environment and realize the secure encryption and transmission of device lifecycle information, and the dynamic protection system integrates BIM model and chaotic encryption mechanism; S3: Build an intelligent diagnosis model to extract features and identify faults of the device running state, provide core data support for follow-up, and form linkage with S2, the intelligent diagnosis model includes attention mechanism and twin network; S4: According to the fault identification result of S3 and based on digital twin technology, a coupling analysis model of energy flow and fault chain is constructed to generate intelligent operation and maintenance strategy, and specific and intelligent guidance scheme is provided for device operation and maintenance to optimize operation and maintenance efficiency and effect; S5. Combine the results of S4 intelligent operation and maintenance strategy and the results involved in the process of S3 feature extraction and fault identification of device running state, and form a data-driven decision support system through edge computing and knowledge graph fusion.

[0006] Preferably, S1 includes the following steps: S11: Composite configuration of PZT-5H piezoelectric ceramic sheet and TEC1-12706 thermoelectric module, connected in series or parallel, and attached to the outer wall of the pipeline, which can generate 8.2mW power under the condition of 30Hz vibration +20℃ temperature difference, and store to 1F super capacitor through 92% efficiency DC-DC converter; S12: When the super capacitor voltage is greater than or equal to 3.3V, the MSP430MCU drives the sensor to enter 1kHz full sampling mode; When the voltage is less than 2.5V, it automatically switches to 10 minutes / once sleep wake-up mode, and preferentially collects core fault characteristic parameters including but not limited to vibration acceleration and bearing temperature; S13: The deployment position of the energy collection module is determined by the digital twin technology simulation of S4, including but not limited to increasing 3 groups of piezoelectric ceramic arrays in the area with the highest vibration energy density near the water pump impeller through CFD simulation, and improving the energy capture efficiency by 30%; The data collected in S1 includes but is not limited to energy data, vibration energy fluctuation data.

[0007] Preferably, the dynamic protection system integrating BIM model and chaotic encryption mechanism in S2 includes the following steps: S21: Generate chaotic key stream through Lorenz mapping, and perform bitwise XOR encryption processing on the data of the device in BIM model including but not limited to geometric coordinates, material properties and running parameters, and the encryption period is associated with the entropy value of the device running state, which is 5 minutes in normal operation and automatically shortened to 100ms in fault; S22: Introduce chaotic disturbance factor during S21 data transmission process, through formula dynamically obfuscating data parameters, wherein an adaptive adjustment coefficient of 0.01-0.1, a real-time chaotic sequence generated based on time series, represents the update amount of the parameter During the training process, the model optimizes the performance by constantly updating represents the gradient of the parameter , that is, the partial derivative of the loss function with respect to S23: A secondary obfuscation step for transmitting data: superimpose the encrypted BIM model data with the real-time generated chaotic sequence for obfuscation, and the receiving end performs double decryption through the corresponding chaotic key stream, so that the decryption error rate is ≤0.01%.

[0008] Preferably, the specific construction steps in S3 are as follows: S31: Based on the topological connection relationship of the equipment in the BIM model, a spatial attention module is constructed, and the sensor data not limited to vibration and temperature is given differentiated weights, the sensor weight near the key components is improved by 20%-30%, and the feature enhancement in the spatial dimension is realized; S32: A time attention module is constructed using a self-attention mechanism, the attention scores of each time step in the sequence are calculated, the equipment transient fault features are focused, the feature extraction efficiency is improved by more than 40% compared with traditional time series models, and the delay problem of traditional LSTM in capturing transient faults is solved; S33: A device health state twin network is constructed, ResNet-18 is used as the backbone network, and the feature difference degree is calculated through a triplet loss function to realize fault recognition and generate a diagnosis result, wherein the sample feature vector dimension is 128, and 3000 groups of samples need to be trained.

[0009] Preferably, the specific construction steps in S3 are as follows: S31: Based on the topological connection relationship of the equipment in the BIM model, a spatial attention module is constructed, and the sensor data not limited to vibration and temperature is given differentiated weights, the sensor weight near the key components is improved by 20%-30%, and the feature enhancement in the spatial dimension is realized; S32: A time attention module is constructed using a self-attention mechanism, the attention scores of each time step in the sequence are calculated, the equipment transient fault features are focused, the feature extraction efficiency is improved by more than 40% compared with traditional time series models, and the delay problem of traditional LSTM in capturing transient faults is solved; ​​S33: Constructing a device health state twin network, taking ResNet-18 as the backbone network, and using a triplet loss function The feature difference degree is calculated, and the fault recognition is realized by calculating the feature difference degree, so as to generate a diagnosis result, wherein the sample feature vector dimension is 128, and 3000 groups of samples need to be trained.

[0010] Preferably, the mechanical and electrical equipment information management and control method based on BIM technology comprises the following steps: S331: Collecting ≥3000 groups of multi-dimensional data of normal / fault state of equipment, including but not limited to vibration energy fluctuation data, and dividing the training set, the validation set and the test set in the proportion of 8:1:1; S332: Setting the batch_size as 32-64, the initial learning rate as 0.001, and using the cosine annealing strategy to iteratively train for more than 500 rounds until the test set accuracy is ≥98%.

[0011] Preferably, S3 further comprises S34: Fault early warning linkage mechanism: when the super capacitor voltage is less than 2.5V for 30 minutes, the sampling frequency is automatically reduced to 50Hz and the key parameters are preferentially collected.

[0012] Preferably, S4 specifically comprises: S41: Using ANSYSTwinBuilder to construct a device digital twin, importing CFD simulation data to realize high-precision mapping, and building an energy metabolism network, which is used to simulate and analyze the energy generation, conversion, transmission and consumption metabolism process of the equipment during operation, including vibration energy conversion and vibration energy fluctuation analysis; S42: Establishing an association rule between energy parameters and fault modes through machine learning, wherein the energy parameters are determined based on the energy data collected in S1, and the association rule includes the association between vibration energy fluctuation and fault mode; S43: Using the digital twin to simulate the energy impact of different operation strategies, and predictively adjusting the air conditioning chilled water flow to increase the TEG module temperature difference by 5℃, which corresponds to an increase of 12% in power generation, thereby optimizing the equipment operation parameters.

[0013] Preferably, S4 further comprises a multi-level early warning threshold setting, specifically as follows: First level warning: piezoelectric power generation is less than 20% of the average value for 30 minutes, and the temperature difference power generation decreases by 15%, corresponding to impeller cavitation failure, with an accuracy of ≥91%; Second level warning: vibration energy fluctuation is more than 30% of the average value, and radio frequency energy decreases by 25%, corresponding to bearing severe wear, with a response time of ≤50ms; In addition, the twin network output of S33 directly drives the digital twin simulation of S4, when the feature difference calculated by the model exceeds the threshold, the ANSY TwinBuilder in S41 is automatically triggered to perform fault mode simulation, forming a closed-loop process of anomaly detection-virtual verification-strategy generation, the energy data of S1 is directly used as the input parameter of the energy metabolism network of S41, when the vibration energy fluctuation collected by S331 exceeds 30% of the average value and the thermoelectric power generation decreases by 25%, S42 automatically marks as a potential fault signal of compressor bearing wear, triggering a secondary warning process.

[0014] Preferably, the specific steps of S5 are as follows: S51: Real-time filtering and abnormal feature extraction of the original data collected by S1 through the edge computing gateway, only the key feature parameters are uploaded to the cloud, thereby reducing the data transmission volume by 50%; S52: Construct a knowledge graph containing device components, energy parameters and fault modes through Neo4j, and obtain the associated path of motor cooling deficiency-TEG temperature difference reduction-bearing lubrication failure through GNN algorithm analysis, to provide prior knowledge support for the diagnostic model of S3, and form a cyclic iteration mechanism of edge preprocessing-cloud knowledge sedimentation-edge decision optimization; The simulation results of S4 and the knowledge graph of S5 form data interaction: the energy parameters and the associated rules of fault modes mined by S42 are imported into the Neo4j graph database, and the associated nodes of energy parameters-fault types are updated, so that the GNN algorithm analysis of S52 can find deeper fault chains.

[0015] (Three) beneficial effects The application provides an electromechanical equipment information management and control method based on BIM technology. Has the following beneficial effects: (1) The electromechanical equipment information management and control method based on BIM technology in use, through the deployment of multi-modal energy collection module to constitute a self-powered sensor network, adopts the composite configuration of PZT-5H piezoelectric ceramic sheet and TEC1-12706 thermoelectric power generation module, which is attached to the outer wall of the pipeline through series or parallel connection, and can generate stable power under certain working conditions and store in super capacitor. According to the super capacitor voltage, the sensor sampling mode is automatically switched to ensure that the equipment data can be effectively collected under different energy states, without complex wiring, and is not affected by power supply interruption, realizing continuous collection of equipment data, improving the stability and reliability of data collection.

[0016] (2) The BIM technology-based mechanical and electrical equipment information management and control method, when in use, generates a chaotic key stream through Lorenz mapping to encrypt the BIM model data, introduces a chaotic disturbance factor in the transmission process for dynamic confusion and secondary confusion, and the receiving end adopts a pipeline decryption architecture and a redundancy check mechanism. The encryption period is associated with the device operating state entropy value, so the encryption period is automatically shortened when a fault occurs, and the encryption strategy can be dynamically adjusted according to the device operating state. In the face of complex network attacks, the security of information transmission is greatly improved, and the safety of device lifecycle information is effectively ensured.

[0017] (3) The BIM technology-based mechanical and electrical equipment information management and control method, when in use, constructs a spatial attention module and a time attention module to give differentiated weights to sensor data, focus on device transient fault features, and improve feature extraction efficiency. A device health state twin network is constructed, and high-precision fault recognition is achieved through a large number of sample training. At the same time, based on digital twin technology, a coupling analysis model of energy flow and fault chain is constructed to generate intelligent operation and maintenance strategies, and through edge computing and knowledge graph fusion, a data-driven decision support system is formed. It can quickly and accurately identify device faults, predictively adjust device operating parameters, optimize operation and maintenance strategies, reduce energy waste and operation and maintenance costs, and improve the operation and maintenance efficiency and management level of mechanical and electrical equipment. BRIEF DESCRIPTION OF DRAWINGS

[0018] Figure 1 is a flowchart of the method of the present application; Figure 2 is a schematic diagram of the sampling mode switching mechanism of the present application. DETAILED DESCRIPTION

[0019] The technical solutions in the embodiments of the present application will be described clearly and completely below with reference to the drawings in the embodiments of the present application. Obviously, the described embodiments are only part of the embodiments of the present application, not all. Based on the embodiments in the present application, all other embodiments obtained by those of ordinary skill in the art without creative labor fall within the scope of the present application.

[0020] Please refer to Figure 1 Figure 2 The present application provides a BIM technology-based mechanical and electrical equipment information management and control method, comprising the following steps: S1: Deploy a multi-modal energy collection module to form a self-powered sensor network to realize continuous collection of device data and provide a source of raw data, and the specific steps are as follows: ​S11: It adopts a composite configuration of PZT-5H piezoelectric ceramic sheet and TEC1-12706 thermoelectric power generation module. The connection method is series or parallel and it is attached to the outer wall of the pipe. Under the condition of 30Hz vibration + 20℃ temperature difference, it can generate 8.2mW of power, which is stored in a 1F supercapacitor through a 92% efficiency DC-DC converter. S12: When the supercapacitor voltage is ≥3.3V, the MSP430MCU drives the sensor to enter the 1kHz full sampling mode; when the voltage is <2.5V, it automatically switches to the sleep wake-up mode every 10 minutes, prioritizing the acquisition of core fault characteristic parameters such as vibration acceleration and bearing temperature. S13: The deployment location of the energy harvesting module is determined by the digital twin simulation in S4, including but not limited to adding 3 sets of piezoelectric ceramic arrays in the area with the highest vibration energy density near the pump impeller through CFD simulation to improve the energy capture efficiency by 30%. The CFD simulation uses ANSYS Fluent 2024 R1 software, and the collected data includes but is not limited to energy data and vibration energy fluctuation data. S2: A dynamic protection system is constructed based on the data collected in S1. To provide a secure data environment and achieve secure encryption and transmission of equipment information throughout its entire lifecycle, the dynamic protection system integrates BIM models and chaotic encryption mechanisms. Specifically, the dynamic protection system integrating BIM models and chaotic encryption mechanisms includes the following steps: S21: Generate a chaotic key stream through Lorentz mapping, and perform bitwise XOR encryption on the data of the equipment in the BIM model, including but not limited to geometric coordinates, material properties and operating parameters. The encryption period is related to the entropy value of the equipment's operating status. The period is 5 minutes during normal operation and is automatically shortened to 100ms during failure. S22: A chaotic perturbation factor is introduced during the data transmission process in S21, using the formula... Dynamically obfuscate the data parameters, where The adaptive adjustment coefficient is 0.01-0.1. This is a real-time chaotic sequence generated based on time series data. Indicates parameters The update frequency. During training, the model is continuously updated. To optimize performance, Indicates parameters The gradient of the loss function, that is, the gradient of the loss function. The partial derivatives; S23: Secondary confusion step of transmitting data: superimpose the encrypted BIM model data with the real-time generated chaotic sequence for confusion, and the receiving end performs double decryption through the corresponding chaotic key stream, so that the decryption error rate is ≤0.01%, specifically, the receiving end adopts a pipeline decryption architecture: first, the FPGA performs parallel calculation of the inverse operation of the chaotic key stream to complete the first decryption, and then a special confusion elimination module is used to perform reverse processing on the superimposed confused data. In addition, in order to ensure decryption accuracy, a redundancy check mechanism is set, and a 16-bit CRC check code is embedded in the encrypted data. If the check fails after decryption, automatic retransmission is requested.

[0021] S3: Build an intelligent diagnosis model to extract features and identify faults of the device running state, mainly to provide core data support for the follow-up, and form a linkage with S2. The intelligent diagnosis model includes attention mechanism and twin network, and the specific construction steps are as follows: S31: Based on the topological connection relationship of the equipment in the BIM model, a spatial attention module is constructed to give differentiated weights to sensor data such as vibration and temperature. The weight of the sensor close to the key component is increased by 20%-30%, realizing feature enhancement in the spatial dimension. Let the initial weight of a certain sensor be , the weight close to the key component is , then: Wherein, is the weight increase coefficient, the value range is 0.2-0.3.

[0022] S32: A time attention module is constructed using self-attention mechanism, which focuses on the transient fault features of the equipment by calculating the attention scores of each time step in the sequence. The feature extraction efficiency is improved by more than 40% compared with traditional time series models, solving the delay problem of traditional LSTM in capturing transient faults. The specific steps are as follows: 1. Map the input sequence to the query vector , the key vector and the value vector , then , wherein , , are weight matrices; 2. Calculate the attention score , the formula is: Wherein, is the dimension of the key vector.

[0023] S33: Construct a device health state twin network, using ResNet-18 as the backbone network, and using a three-tuple loss function The feature difference is calculated, where the sample feature vector dimension is 128, and 3000 groups of samples are trained, and the specific training process is as follows: S331: Collect multi-dimensional data of ≥3000 groups of equipment normal / failure states, and divide the training set, validation set and test set according to the ratio of 8:1:1; S332: Set batch_size to 32-64, initial learning rate to 0.001, and use the cosine annealing strategy to iterate training for more than 500 rounds until the test set accuracy is ≥98%; In addition, S3 also includes S35: Fault early warning linkage mechanism: when the super capacitor voltage is less than 2.5V for 30 minutes, the sampling frequency is automatically reduced to 50Hz and the key parameters are preferentially collected.

[0024] Further, the encryption logic of S2 and the diagnosis model of S3 form a closed-loop linkage. When S2 identifies an equipment anomaly, it automatically triggers the encryption period shortening mechanism in S21, and at the same time, the abnormal feature parameters are included in the entropy value calculation factor of the chaotic sequence generation, forming a "state recognition-encryption enhancement" adaptive security link. The chaotic sequence generation uses an improved Logistic mapping, and the specific formula is as follows: wherein, 3.99, is the entropy value of the equipment operating state, and the initial value is the normalized value of the current timestamp in milliseconds, represents the state variable of the system at step , usually taking a value in a certain interval, represents the state of the system at step , which is determined by the state of the previous step, and embodies the "recursion" characteristic.

[0025] S4: According to the diagnosis results of S3 and based on digital twin technology, a coupling analysis model of energy flow and fault chain is constructed to generate intelligent operation and maintenance strategies. The specific process is as follows: S41: Use ANSYSTwinBuilder to construct the device digital twin, import CFD simulation data to realize high-precision mapping, and build an energy metabolism network. This network is used to simulate and analyze the energy production, conversion, transmission and consumption of the device during operation, including the conversion of vibration energy and the analysis of vibration energy fluctuation; S42: Establishing the association rules between the energy parameters and the failure modes through machine learning, wherein the energy parameters are determined based on the energy data collected in S1, and the association rules include the association between the vibration energy fluctuation and the failure modes; S43: Using the digital twin to simulate the energy impact of different operation strategies, and predictively adjusting the air conditioning chilled water flow to make the TEG module temperature difference increase by 5°C, corresponding to an increase of 12% in power generation, thereby optimizing the equipment operation parameters; Further description, S4 also includes multi-level early warning threshold setting, specifically as follows: First level warning: piezoelectric power generation for 30 consecutive minutes < 20% of the average value and temperature difference power generation decreases by 15%, corresponding to impeller cavitation failure, accuracy ≥ 91%; Second level warning: vibration energy fluctuation exceeds the average value by 30% and radio frequency energy decreases by 25%, corresponding to severe bearing wear, response time ≤ 50ms; In addition, the twin network output of S33 directly drives the digital twin simulation of S4. When the feature difference calculated by the model exceeds the threshold, ANSYSTwinBuilder in S41 is automatically triggered for failure mode simulation, forming a closed-loop process of anomaly detection-virtual verification-strategy generation. The energy data collected by S1 is directly used as the input parameter of the energy metabolism network in S41. When S331 collects vibration energy fluctuation exceeding the average value by 30% and temperature difference power generation decreasing by 25%, S42 automatically marks it as a potential failure signal of compressor bearing wear, triggering the second level warning process.

[0026] S5. Through edge computing and knowledge graph fusion, a data-driven decision support system is formed, and the specific steps are as follows: S51: Through the edge computing gateway, the original data collected by S1 is filtered and abnormal features are extracted in real time, and only the key feature parameters are uploaded to the cloud, thereby reducing the data transmission volume by 50%; S52: Through Neo4j, a knowledge graph containing device components, energy parameters and failure modes is constructed, and the association path of motor cooling deficiency-TEG temperature difference decrease-bearing lubrication failure is obtained through GNN algorithm analysis, providing prior knowledge support for the diagnosis model of S3, and forming a cyclic iteration mechanism of edge preprocessing-cloud knowledge sedimentation-edge decision optimization; The simulation results of S4 and the knowledge graph of S5 form data interaction: the association rules between the energy parameters and the failure modes mined by S42 are imported into the Neo4j graph database, and the association nodes of the energy parameters-failure types are updated, so that the GNN algorithm analysis of S52 can find deeper failure chains.

[0027] While embodiments of the application have been shown and described, it is to be understood that the embodiments described are merely exemplary of the principles and application of the present application. Numerous modifications and adaptions can be effected without departing from the spirit and scope of the present application, which is not limited to the exact construction and arrangement described. It is intended, therefore, to cover all modifications and adaptions that fall within the scope of the claims and their equivalents.

Claims

1. A BIM technology-based mechanical and electrical equipment information management and control method, characterized in that, Comprise the following steps: S1: Deploy a multi-modal energy harvesting module to form a self-powered sensor network, realize continuous acquisition of device data, and provide a source of raw data; S2: Construct a dynamic protection system according to the data collected in S1 to provide a secure data environment and realize the secure encryption and transmission of device lifecycle information, the dynamic protection system combines BIM models and chaotic encryption mechanisms; S3: Build an intelligent diagnostic model to extract features and identify faults of the device operating state, provide core data support for follow-up, and form a linkage with S2, the intelligent diagnostic model includes an attention mechanism and a twin network; S4: According to the fault identification results of S3 and based on digital twin technology, build a coupling analysis model of energy flow and fault chain to generate intelligent operation and maintenance strategies and provide specific and intelligent guidance schemes for device operation and maintenance to optimize operation and maintenance efficiency and effectiveness; S5. Combine the intelligent operation and maintenance strategies of S4 and the feature extraction and fault identification results of S3, and form a data-driven decision support system through edge computing and knowledge graph fusion. 2.The mechanical and electrical equipment information management and control method based on BIM technology according to claim 1, characterized in that: The specific steps of S1 include: S11: Composite configuration of PZT-5H piezoelectric ceramic sheet and TEC1-12706 thermoelectric generator module, connection mode is series or parallel, and attached to the outer wall of the pipeline, can generate 8.2mW power under the condition of 30Hz vibration +20℃ temperature difference, through 92% efficiency DC-DC converter to 1F super capacitor; S12: When the super capacitor voltage is greater than or equal to 3.3V, the MSP430MCU drives the sensor to enter the 1kHz full sampling mode; When the voltage is less than 2.5V, it automatically switches to the sleep wake-up mode of 10 minutes / time, and preferentially collects core fault characteristic parameters including but not limited to vibration acceleration and bearing temperature; S13: The deployment position of the energy harvesting module is determined by digital twin technology simulation of S4, including but not limited to increasing 3 groups of piezoelectric ceramic arrays in the area with the highest vibration energy density near the water pump impeller through CFD simulation, and improving the energy capture efficiency by 30%; The device data collected in S1 includes but is not limited to energy data and vibration energy fluctuation data. 3.The mechanical and electrical equipment information management and control method based on BIM technology according to claim 1, characterized in that: The dynamic protection system in S2 that combines BIM models and chaotic encryption mechanisms includes the following steps: S21: Generate a chaotic key stream through Lorenz mapping, and perform bitwise XOR encryption processing on the data of the device in the BIM model including but not limited to geometric coordinates, material properties and operating parameters, the encryption period is associated with the device operating state entropy value, and the period is 5 minutes in normal operation and automatically shortened to 100ms in fault; S22: Introduce a chaotic disturbance factor during the data transmission process of S21, through the formula dynamically obfuscating data parameters, wherein an adaptive adjustment factor, is a real-time chaotic sequence generated based on time series, denotes the gradient of the parameters , that is, the partial derivative of the loss function with respect to to optimize performance, denotes the gradient of the parameters , that is, the partial derivative of the loss function with respect to . S23: Secondary confusion step of transmission data: superimpose and confuse the encrypted BIM model data with the real-time generated chaotic sequence, and the receiving end performs double decryption through the corresponding chaotic key stream, so that the decryption error rate is less than or equal to 0.01%.

4. The mechanical and electrical equipment information management and control method based on BIM technology according to claim 3, characterized in that: The specific construction steps in S3 are as follows: S31: Construct a spatial attention module based on the topological connection relationship of the equipment in the BIM model, and give different weights to sensor data such as vibration and temperature. The weight of the sensor close to the key component is increased by 20%-30%, realizing feature enhancement in the spatial dimension; S32: Construct a time attention module using a self-attention mechanism. By calculating the attention score of each time step in the sequence, focus on the transient fault features of the equipment. The feature extraction efficiency is improved by more than 40% compared with traditional time sequence models, solving the delay problem of traditional LSTM in capturing transient faults; S33: Construct a device health state twin network, taking ResNet-18 as a backbone network, and passing through a triple loss function Calculate the feature difference degree, realize fault recognition through calculating the feature difference degree, and thus generate a diagnosis result, wherein The sample feature vector dimension is 128, and 3000 groups of samples need to be trained.

5. The mechanical and electrical equipment information management and control method based on BIM technology according to claim 4, characterized in that: The chaotic encryption mechanism of S2 and the diagnostic model of S3 form a closed-loop linkage. When S2 identifies an equipment anomaly, it automatically triggers the encryption period shortening mechanism in S21, and at the same time, the abnormal feature parameters are included in the entropy calculation factor of the chaotic sequence generation, forming a state recognition-encryption reinforcement adaptive security link.

6. The mechanical and electrical equipment information management and control method based on BIM technology according to claim 4, characterized in that: The training process of S33 is as follows: S331: Collect ≥3000 groups of multi-dimensional data of equipment in normal / fault state, including but not limited to vibration energy fluctuation data, and divide the training set, validation set and test set in the ratio of 8:1:1; S332: Set batch_size to 32-64, initial learning rate to 0.001, and use cosine annealing strategy to iterate training for more than 500 rounds until the test set accuracy is ≥98%.

7. The mechanical and electrical equipment information management and control method based on BIM technology according to claim 4, characterized in that: S3 also includes S34: Fault warning linkage mechanism: When the super capacitor voltage is less than 2.5V for 30 minutes, the sampling frequency is automatically reduced to 50Hz and the key parameters are preferentially collected.

8. The mechanical and electrical equipment information management and control method based on BIM technology according to claim 6, characterized in that: S4 specifically includes: S41: Use ANSYSTwinBuilder to construct the digital twin of the equipment, import CFD simulation data to realize high-precision mapping, and build an energy metabolism network. This network is used to simulate and analyze the metabolism process of energy generation, conversion, transmission and consumption of the equipment during operation, including vibration energy conversion and vibration energy fluctuation analysis; S42: Establish the association rules between energy parameters and fault modes through machine learning, where the energy parameters are determined based on the energy data collected by S1, and the association rules include the association between vibration energy fluctuation and fault mode; S43: Use the digital twin to simulate the energy impact of different operation strategies, and predictively adjust the air conditioning chilled water flow to increase the TEG module temperature difference by 5°C, corresponding to an increase in power generation by 12%, thereby optimizing the equipment operation parameters.

9. The mechanical and electrical equipment information management and control method based on BIM technology according to claim 8, characterized in that: S4 also includes multi-level warning threshold setting, specifically as follows: Primary warning: Piezoelectric power generation is less than 20% of the average value for 30 minutes, and the temperature difference power generation decreases by 15%, corresponding to impeller cavitation failure, with an accuracy of ≥91%; Secondary warning: Vibration energy fluctuation is more than 30% of the average value, and radio frequency energy decreases by 25%, corresponding to bearing severe wear, with a response time of ≤50ms; In addition, the twin network output of S33 directly drives the digital twin simulation of S4. When the characteristic difference calculated by the model exceeds the threshold, the ANSYSTwinBuilder in S41 is automatically triggered to perform fault mode simulation, forming a closed-loop process of anomaly detection-virtual verification-strategy generation. The energy data of S1 is directly used as the input parameter of the energy metabolism network in S41. When S331 collects vibration energy fluctuations exceeding the mean value by 30% and the temperature difference power generation decreases by 25%, S42 automatically marks it as a potential fault signal of compressor bearing wear, triggering a secondary warning process.

10. The mechanical and electrical equipment information management and control method based on BIM technology according to claim 8, characterized in that: The specific steps of S5 are as follows: S51: Real-time filtering and abnormal feature extraction of the original data collected by S1 through the edge computing gateway, only uploading key feature parameters to the cloud, thereby reducing the data transmission volume by 50%; S52: Constructing a knowledge graph containing device components, energy parameters, and fault modes through Neo4j, and analyzing the associated path of motor cooling deficiency-TEG temperature difference reduction-bearing lubrication failure through GNN algorithm to provide prior knowledge support for the diagnostic model of S3, forming a cyclic iteration mechanism of edge preprocessing-cloud knowledge sedimentation-edge decision optimization; The simulation results of S4 and the knowledge graph of S5 form data interaction: the correlation rules of energy parameters and fault modes mined by S42 are imported into the Neo4j graph database, updating the correlation nodes of energy parameters-fault types, so that the GNN algorithm analysis of S52 can find deeper fault chains.

Images