Industrial cloud platform compressor air system management and control system based on Internet of Things gateway

Abstract

The invention discloses an industrial cloud platform compressor air system management and control system based on an Internet of Things gateway. The data acquisition unit acquires operation data of the compressor through a sensor, and transmits the operation data to the Internet of Things gateway unit after preprocessing the operation data. And the Internet of Things gateway unit performs end-to-end encryption processing, equipment identity verification and dynamic access control on the data. And after receiving and decrypting the data, the industrial cloud platform digital twin unit constructs and updates a digital twin model based on real-time data, extracts key performance indexes by using a stream computing technology, and stores the key performance indexes in a time sequence database to support analysis and optimization. And the intelligent diagnosis optimization unit performs fault diagnosis in combination with the key performance indexes and high-order spectrum analysis, and optimizes operation parameters of the compressor through a reinforcement learning algorithm. The distributed autonomous control unit adjusts a real-time operation state based on optimization parameters, realizes load adaptive balance and energy efficiency optimization, starts a local autonomous control strategy when communication is interrupted, and ensures safe and efficient operation of the system.

Description

Technical Field

[0001] The present invention relates to the technical field of control and regulation systems, and in particular to a management and control system for an industrial cloud platform compressor air system based on an Internet of Things gateway. Background Art

[0002] In the context of existing technologies, the monitoring of traditional compressor air systems relies on simple sensors and centralized monitoring methods. Although they can achieve basic operating data collection and analysis, they are significantly insufficient in terms of real-time response and intelligence. The gas demand at the workshop end may change instantly due to drastic fluctuations in operating conditions, and it is difficult for traditional systems to adjust the operating status of the unit in a short period of time to cope with demand fluctuations, which may lead to problems such as insufficient gas supply or overload operation. In addition, existing systems mostly use basic data analysis methods, which cannot fully utilize real-time and historical data to predict the possibility of failures, and it is difficult to provide operators with specific warnings or solution suggestions, resulting in potential risks not being dealt with in a timely manner.

[0003] The introduction of industrial Internet of Things technology provides new capabilities for remote monitoring and data integration for the management of compressor systems, but the current security issues of data transmission and storage remain unresolved. Sensitive operating data may be leaked or tampered with due to insufficient security mechanisms during uploading to the cloud platform. Especially in industrial environments, this risk may have a serious impact on factory operations and production secrets. In addition, although digital twin technology has been tried to be applied to compressor modeling, the existing system lacks efficient data processing capabilities and intelligent optimization algorithms, making it difficult to dynamically adjust operating parameters based on diagnostic results and performance indicators, thus affecting the overall efficiency and long-term reliability of the system.

[0004] The ability to respond in emergencies is also a shortcoming of existing technologies. In the event of communication interruption or network loss, most existing systems rely on centralized control, lack distributed autonomy, and cannot provide effective localized control strategies in standalone mode. This limitation may result in the inability to effectively adjust the compressor operating status in an emergency, seriously affecting the stability and safety of production.

[0005] Therefore, the present invention has developed a new type of compressor air system control system based on the Internet of Things gateway and industrial cloud platform, aiming to solve these core problems. By quickly responding to fluctuations in air consumption at the workshop end, predicting faults based on data trends, providing intelligent warnings and solutions, and realizing dynamic visualization, the present invention not only improves the intelligence level of the system, but also enhances its reliability and safety in emergency situations. Summary of the invention

[0006] The present application provides a management and control system for an industrial cloud platform compressor air system based on an Internet of Things gateway to improve the operating efficiency of the industrial compressor air system.

[0007] The present application provides a control system for an industrial cloud platform compressor air system based on an Internet of Things gateway, including:

[0008] A data acquisition unit is used to collect the operating data of the compressor in real time through intelligent sensors configured at the air inlet end, the exhaust end and the body of the compressor, wherein the operating data includes temperature, pressure, vibration and current; pre-process the collected operating data, and transmit the pre-processed operating data to the Internet of Things gateway unit;

[0009] The IoT gateway unit is used to receive the pre-processed operating data transmitted by the data acquisition unit; perform end-to-end encryption on the received data; verify the device identity of the data source and perform dynamic access control to ensure the security of data transmission; and transmit the encrypted and verified operating data to the digital twin unit of the industrial cloud platform;

[0010] The digital twin unit of the industrial cloud platform is used to receive the encrypted operation data transmitted by the IoT gateway unit; after decrypting and verifying the operation data, the digital twin model of the compressor system is constructed and updated in real time based on the received data, and the digital twin model is used to synchronously map the physical system and information system of the compressor; the real-time operation data in the digital twin model is analyzed by using stream computing technology to extract key performance indicator data, and the key performance indicator data includes energy efficiency ratio, reliability and health; the extracted key performance indicator data is synchronously stored with the original operation data in the time series database to support historical trend analysis, fast retrieval and data modeling optimization; the extracted key performance indicator data is transmitted to the intelligent diagnosis optimization unit;

[0011] An intelligent diagnosis and optimization unit is used to receive key performance indicators transmitted by the digital twin unit of the industrial cloud platform; based on the key performance indicators, high-order spectrum analysis is used to perform fault diagnosis on the operating status of the compressor to generate fault diagnosis results; combining the fault diagnosis results with the key performance indicators, a reinforcement learning algorithm is used to optimize the operating parameters of the compressor to generate optimized operating parameters, wherein the optimized operating parameters include a pressure setting value, a load distribution strategy, and an energy efficiency optimization plan;

[0012] The distributed autonomous control unit is used to receive the optimized operating parameters and fault diagnosis results fed back by the intelligent diagnosis and optimization unit; adjust the real-time operating status of the compressor according to the optimized operating parameters to achieve load adaptive balancing and energy efficiency optimization; in the event of communication interruption, start the local autonomous control strategy according to the fault diagnosis results and execute emergency control under emergency conditions.

[0013] Furthermore, the data acquisition unit is specifically used for:

[0014] Aggregate the collected data into local data cache through sensor nodes to form small batch data blocks;

[0015] After the data block is formed, the data block is sent to the IoT gateway unit at fixed time intervals using a timing transmission protocol;

[0016] During the transmission process, data verification information is added, including acquisition timestamp, sensor identification and data summary verification value, to ensure the integrity and correctness of the data during transmission.

[0017] Furthermore, the data acquisition unit includes a data preprocessing module, and the data preprocessing module is used to:

[0018] Data denoising to remove environmental interference signals, including outlier detection based on sensor signal consistency;

[0019] Data integrity check: by verifying the continuity and validity of the collected signal, the abnormal data segments that do not meet the requirements are eliminated;

[0020] Data standardization converts the operating data into a uniform numerical range for subsequent processing and analysis.

[0021] Furthermore, the Internet of Things gateway unit includes the following modules:

[0022] A data receiving module is used to receive the pre-processed operation data from the data acquisition unit and perform consistency check on the timestamp and source device identification of the data;

[0023] An encryption module, used to perform end-to-end encryption processing on the verified data, wherein the encryption process dynamically generates a key and synchronizes the key with the digital twin unit of the industrial cloud platform;

[0024] Authentication module, used to verify the identity of the data source device, including device identification comparison and historical trust record check;

[0025] The dynamic access control module is used to assign or adjust access rights based on the verification results of the device identity and monitor abnormal behavior during data interaction in real time.

[0026] This application has the following beneficial technical effects:

[0027] (1) The intelligent sensors of the data acquisition unit collect the operating data of the compressor in real time, and the digital twin model is updated in real time using the digital twin unit of the industrial cloud platform. This can efficiently synchronize the physical system and the information system, ensure the dynamic reflection and accurate analysis of the operating status, and significantly improve the real-time and accuracy of system monitoring. (2) The IoT gateway unit ensures the security of data during transmission and storage through end-to-end encryption processing, device identity authentication and dynamic access control, effectively preventing data leakage and illegal access, and providing security for the reliable operation of the industrial compressor system. (3) Combined with the intelligent diagnosis and optimization unit, fault diagnosis is achieved through high-order spectrum analysis, and the compressor operating parameters are intelligently optimized using reinforcement learning algorithms, significantly improving the operating efficiency and reliability of the system. (4) In the event of an emergency when communication is interrupted, the distributed autonomous control unit can start the local autonomous control strategy based on the fault diagnosis results to maintain the basic operation of the system and perform emergency processing, effectively improving the stability and adaptability of the system in the face of emergencies. BRIEF DESCRIPTION OF THE DRAWINGS

[0028] Figure 1 It is a schematic diagram of a control system of an industrial cloud platform compressor air system based on an Internet of Things gateway provided in the first embodiment of the present application. DETAILED DESCRIPTION

[0029] Many specific details are described in the following description to facilitate a full understanding of the present application. However, the present application can be implemented in many other ways than those described herein, and those skilled in the art can make similar generalizations without violating the connotation of the present application, so the present application is not limited by the specific implementation disclosed below.

[0030] The first embodiment of the present application provides a control system for the compressor air system of an industrial cloud platform based on an Internet of Things gateway. Figure 1 , which is a schematic diagram of the first embodiment of the present application. Figure 1 The first embodiment of the present application provides a control system for an industrial cloud platform compressor air system based on an Internet of Things gateway, which is described in detail.

[0031] The control system of the compressor air system of the industrial cloud platform based on the Internet of Things gateway includes a data acquisition unit 101, an Internet of Things gateway unit 102, an industrial cloud platform digital twin unit 103, an intelligent diagnosis and optimization unit 104 and a distributed autonomous control unit 105.

[0032] The data acquisition unit 101 is used to collect the operating data of the compressor in real time through intelligent sensors configured at the air inlet end, the exhaust end and the body of the compressor, and the operating data includes temperature, pressure, vibration and current; pre-process the collected operating data, and transmit the pre-processed operating data to the Internet of Things gateway unit.

[0033] First, the intelligent sensor array is the basic component of the data acquisition unit 101. The array includes multiple sensors installed at the intake end, exhaust end and body surface of the compressor. These sensors are used to collect operating parameters such as temperature, pressure, vibration and current. For example, the intake end sensor monitors the temperature and pressure of the air entering the compressor, the exhaust end sensor collects the temperature and pressure changes of the discharged air, the vibration sensor on the body records the mechanical vibration state of the compressor in real time during operation, and the current sensor is used to measure the operating current of the compressor motor to reflect the power consumption.

[0034] Secondly, the data preprocessing module integrated in the data acquisition unit 101 can perform preliminary processing on the collected operation data, including operations such as denoising, integrity checking and format conversion. The denoising function is used to filter out irrelevant signals caused by environmental interference or device jitter, thereby ensuring the accuracy of the data; the integrity check ensures that the collected data is not lost or damaged, such as by adding a check code to check the reliability of the data transmission process; the format conversion function unifies data in various formats into a system standard format, which is convenient for subsequent module processing and analysis.

[0035] Next, the pre-processed data is transmitted from the data acquisition unit (101) to the Internet of Things gateway unit (102) via a high-speed communication interface. In order to adapt to the large amount of real-time data generated during the operation of the compressor, the data acquisition unit (101) adopts a segmented data transmission method to package the real-time collected data in the form of small batch blocks and transmit them to the gateway unit. This segmented transmission method can not only reduce data transmission delay, but also reduce transmission failures caused by network interruption or data packet loss.

[0036] In addition, in order to adapt to the complex requirements of different industrial environments, the data acquisition unit (101) supports multiple working modes, including real-time monitoring mode and timed sampling mode. In the real-time monitoring mode, the sensor collects operating data at the highest frequency to capture emergencies or abnormal conditions. In the timed sampling mode, the system collects data at a preset sampling interval, which is suitable for monitoring the long-term stable operation of the compressor. The user can flexibly configure the sampling mode through the system interface or the remote management platform.

[0037] Finally, in order to improve the reliability and adaptability of the entire system, the data acquisition unit 101 also has a self-checking function. The self-checking function can regularly detect the operating status of the sensor and data preprocessing module. When a hardware failure or data anomaly is detected, the system will generate an alarm message and send it to the cloud platform through the gateway unit to ensure that potential problems can be discovered and handled in a timely manner.

[0038] In summary, the design of the data acquisition unit 101 realizes efficient and accurate collection of the compressor operating status, and optimizes the data quality through the preprocessing module, providing a reliable data foundation for subsequent gateway transmission, cloud analysis and diagnostic optimization.

[0039] Furthermore, the data acquisition unit is specifically used for:

[0040] Aggregate the collected data into local data cache through sensor nodes to form small batch data blocks;

[0041] After the data block is formed, the data block is sent to the IoT gateway unit at fixed time intervals using a timing transmission protocol;

[0042] During the transmission process, data verification information is added, including acquisition timestamp, sensor identification and data summary verification value, to ensure the integrity and correctness of the data during transmission.

[0043] The data acquisition unit first aggregates the collected operation data into the local data cache through the configured sensor nodes to form small batches of data blocks. This data aggregation method can reduce the amount of data transmitted in a single time and improve transmission efficiency. After the data blocks are formed, the timing transmission protocol is used to send the data blocks to the IoT gateway unit at fixed time intervals. The use of the timing transmission protocol ensures the timeliness of data transmission while reducing the possibility of data loss or delay. In addition, during the transmission of the data block, the system will attach verification information including the acquisition timestamp, sensor identification and data summary check value. This additional information allows the receiving end to verify the source, time accuracy and content integrity of the data, thereby further improving the reliability and security of data transmission.

[0044] Furthermore, the data acquisition unit includes a data preprocessing module, and the data preprocessing module is used to:

[0045] Data denoising to remove environmental interference signals, including outlier detection based on sensor signal consistency;

[0046] Data integrity check: by verifying the continuity and validity of the collected signal, the abnormal data segments that do not meet the requirements are eliminated;

[0047] Data standardization converts the operating data into a uniform numerical range for subsequent processing and analysis.

[0048] The data acquisition unit includes a data preprocessing module, which is used to optimize the quality and format of the collected data. First, the data preprocessing module performs data denoising to ensure the accuracy of the collected data by eliminating environmental interference signals, including an outlier detection method based on sensor signal consistency to identify and eliminate possible abnormal signals. Next, the module will perform an integrity check on the data, identify and eliminate abnormal data segments by analyzing the continuity and validity of the signal, and ensure that the data passed to the downstream system meets the basic quality requirements. Finally, the data preprocessing module standardizes the operating data and converts the data into a unified numerical range. The standardized data is easier to analyze and process subsequently, avoiding analysis bias or algorithm adaptation problems caused by differences in data distribution.

[0049] The IoT gateway unit 102 is used to receive the pre-processed operating data transmitted by the data acquisition unit; perform end-to-end encryption on the received data; verify the device identity of the data source and perform dynamic access control to ensure the security of data transmission; and transmit the encrypted and verified operating data to the digital twin unit of the industrial cloud platform.

[0050] The IoT gateway unit 102 is one of the key components, used to bridge the communication between the data acquisition unit and the industrial cloud platform to ensure the security and reliability of data transmission and the effectiveness of data processing. The design and implementation of this unit includes multiple functional modules and specific processes to meet the needs of efficient management and control of industrial compressor air systems.

[0051] The IoT gateway unit first receives the pre-processed operating data transmitted from the data acquisition unit 101 through its communication interface. The operating data includes various parameters such as the temperature, pressure, vibration and current of the compressor, which arrive at the gateway unit in batch or streaming mode. The received data will undergo a preliminary integrity check to ensure that the data packet is not damaged or lost during transmission. The integrity check usually relies on the verification information attached to the data packet, such as a timestamp, data length or hash value.

[0052] To ensure the confidentiality of the data, the IoT gateway unit performs end-to-end encryption on the received operation data. The encryption module adopts a dynamic key management mechanism to share symmetric or asymmetric keys with the industrial cloud platform. The operation data is stored in the gateway's temporary buffer after encryption and is further processed when it is transmitted to the industrial cloud platform. The encryption process ensures that even if the data is intercepted during transmission, it cannot be interpreted. In addition, to enhance transmission efficiency and adapt to different network bandwidths, the gateway unit supports flexible selection of multiple encryption algorithms, including lightweight encryption algorithms to adapt to low-bandwidth scenarios, and high-intensity encryption algorithms to cope with highly sensitive data.

[0053] Device authentication is one of the important functions of the IoT gateway unit. Each data acquisition unit sends device identification information, including device ID, model, and registration authentication information, when it first connects to the gateway. The gateway unit verifies this information through the internal authentication module, including checking the device's historical connection records and matching results with the pre-stored device list. Once an unauthorized device attempts to access, the gateway will immediately reject its connection request and record the event for subsequent review. The device authentication process not only ensures the access of legitimate devices, but also effectively prevents the occurrence of counterfeit devices or malicious attacks.

[0054] Dynamic access control is another important function of the IoT gateway unit. The access control module dynamically assigns permissions based on the device's authentication results and real-time operating conditions. For example, for a normally operating acquisition unit, the gateway allows it to transmit data with the highest priority; for devices with abnormalities or potential failures, the gateway will lower its transmission priority or even suspend its data upload operation to protect the overall performance of the system. The rules of dynamic access control can be adjusted through preset policies or remote configuration by the administrator to adapt to different industrial scenarios.

[0055] After completing data encryption and security verification, the IoT gateway unit transmits the processed data to the industrial cloud platform digital twin unit 103 through its uplink communication interface. During the data transmission process, the gateway attaches metadata, including device ID, timestamp, and data source description, so that the cloud platform can correctly parse the data and build a digital twin model. In order to further ensure the security of data transmission, the gateway supports a two-way authentication protocol with the industrial cloud platform to ensure the integrity and anti-tampering capability of the transmission link.

[0056] Furthermore, the Internet of Things gateway unit includes the following modules:

[0057] A data receiving module is used to receive the pre-processed operation data from the data acquisition unit and perform consistency check on the timestamp and source device identification of the data;

[0058] An encryption module, used to perform end-to-end encryption processing on the verified data, wherein the encryption process dynamically generates a key and synchronizes the key with the digital twin unit of the industrial cloud platform;

[0059] Authentication module, used to verify the identity of the data source device, including device identification comparison and historical trust record check;

[0060] The dynamic access control module is used to assign or adjust access rights based on the verification results of the device identity and monitor abnormal behavior during data interaction in real time.

[0061] As a key communication and security bridge in the system, the IoT gateway unit consists of multiple functional modules that work together to receive, process, encrypt and verify data, ensuring the integrity and security of data transmission while improving the system's reliability and risk resistance.

[0062] The data receiving module is responsible for obtaining pre-processed operating data from the data acquisition unit. The operating data includes parameters such as temperature, pressure, vibration, and current, and is accompanied by a timestamp and source device identification information. During the receiving process, the data receiving module performs a consistency check on each piece of data to ensure the sequence and integrity of the timestamp, and to verify the consistency of the source device identification with the expected sending device. If the check fails, the data will be rejected and the exception handling mechanism will be triggered.

[0063] The encryption module performs end-to-end encryption on the verified operating data to ensure the confidentiality and tamper resistance of the data during transmission. The encryption process adopts a dynamic key generation mechanism to generate an independent encryption key each time the data is interacted. This mechanism can significantly reduce the risk of key leakage. The generation and use of keys are completed by the gateway unit and the industrial cloud platform digital twin unit, and key synchronization is used to ensure the seamless connection of the encryption and decryption process. The design of the encryption module supports the flexible selection of multiple encryption algorithms to meet the security and performance requirements of different scenarios.

[0064] The identity verification module is used to verify the legitimacy of the data source device. The module compares the device identification information with the device list stored in the system to ensure that only authorized devices can access the system. In addition, the module also checks the historical trust record of the device to further prevent potential security risks. For example, if a device has recently experienced multiple abnormal data transmissions, the system will lower its trust level and restrict its permissions. By combining device identification comparison and trust record checking, the identity verification module can effectively prevent illegal device access and data forgery.

[0065] The dynamic access control module assigns appropriate access rights to each device based on the verification results of the identity authentication module. The rules for assigning permissions can be pre-set and dynamically adjusted based on the status and historical behavior of the device. For example, devices in good operating condition can be granted full data access rights, while devices with security risks may be restricted to read-only or temporarily prohibited from access. The dynamic access control module also monitors in real time abnormal behaviors that may occur during data interaction, such as excessive data transmission frequency or unauthorized access attempts. Once an anomaly is detected, the module immediately triggers an alarm and takes response measures, including recording the event, reducing device permissions, or temporarily disconnecting the device.

[0066] Through the collaborative work of these modules, the IoT gateway unit achieves flexible access control while ensuring data security, providing high-quality and reliable operating data for the industrial cloud platform.

[0067] The industrial cloud platform digital twin unit 103 is used to receive the encrypted operating data transmitted by the IoT gateway unit; after decrypting and verifying the operating data, a digital twin model of the compressor system is constructed and updated in real time based on the received data, and the digital twin model is used to synchronously map the physical system and information system of the compressor; the real-time operating data in the digital twin model is analyzed by using stream computing technology, and key performance indicator data is extracted, and the key performance indicator data includes energy efficiency ratio, reliability and health; the extracted key performance indicator data is synchronously stored with the original operating data in a time series database to support historical trend analysis, fast retrieval and data modeling optimization; and the extracted key performance indicator data is transmitted to the intelligent diagnosis optimization unit.

[0068] The industrial cloud platform digital twin unit 103 is one of the core modules, responsible for receiving the encrypted operation data transmitted by the IoT gateway unit, and building and updating the digital twin model of the compressor system in real time based on these data after decryption and verification. The functional realization of this unit depends on the collaborative work of multiple logic modules, including data decryption and verification module, digital twin model construction and update module, stream computing analysis module, and time series database storage module.

[0069] During the data reception process, the digital twin unit of the industrial cloud platform first receives encrypted operating data from the IoT gateway through its communication interface. These data include compressor operating parameters such as temperature, pressure, vibration and current, and are accompanied by device identification information and timestamps. The data decryption and verification module decrypts the received operating data to ensure that the data can be accurately parsed. After decryption, the verification module further checks the integrity of the data and the authenticity of the source. Integrity verification is achieved by checking the transmitted checksum value, and the authenticity of the source depends on the comparison of the device identification information and the matching with the system registered device list. Any abnormal data will be marked and stored in an independent log for system administrators to check.

[0070] After decryption and verification, the digital twin model construction and update module begins to initialize or dynamically update the digital twin model of the compressor based on the received data. The digital twin model is a data-driven virtual model that can reflect the physical state and operating characteristics of the compressor in real time. During the initialization phase, the system generates an initial model framework based on the design parameters of the compressor (such as geometry, rated power, and maximum working pressure). During the operation phase, real-time data drives the dynamic update of the model so that the state parameters in the model (such as temperature field, pressure field, and vibration intensity) can synchronize the changes in the physical compressor. This dynamic mapping mechanism ensures that the digital twin model can always accurately reflect the actual operating status of the physical system.

[0071] The real-time operation data in the digital twin model then enters the stream computing analysis module, which continuously analyzes the data to extract key performance indicators. These performance indicators include energy efficiency ratio, reliability and health. The energy efficiency ratio is calculated by the ratio of input power to output power in the operating parameters; reliability is obtained by dynamic evaluation of vibration characteristics, operating time and load changes; health is scored based on the degree of deviation of parameters such as temperature and pressure, reflecting the overall health status of the system. The stream computing analysis module can complete data processing and indicator calculation in milliseconds to meet the real-time requirements of high-frequency data streams.

[0072] The extracted key performance indicator data will be stored in the time series database together with the original operation data. The structured design of the time series database allows the system to quickly retrieve, analyze trends and optimize modeling of historical data. By analyzing historical trends, the system can identify changes in compressor operating modes and provide a basis for predicting future operating conditions. In addition, this data can also be used to train new optimization algorithms to improve the overall intelligence level of the system.

[0073] Finally, the extracted key performance indicators are transmitted to the intelligent diagnosis and optimization unit through a specific interface, providing a direct basis for fault diagnosis and optimization of operating parameters. The design of the digital twin unit ensures high efficiency and high accuracy in the entire process from data reception to processing, analysis and output.

[0074] Furthermore, the industrial cloud platform digital twin unit constructs and updates the digital twin model of the compressor system in real time based on the received data, including the following steps:

[0075] After receiving the encrypted operation data transmitted from the IoT gateway unit, decrypt the data and verify the integrity, source legitimacy and time synchronization of the data;

[0076] When the system is started or data is first received, an initial digital twin model is established based on the physical characteristics and structural parameters of the compressor, the model including basic information on geometric structure, operating status and dynamic parameters;

[0077] Mapping the decrypted and verified real-time operation data to the dynamic parameters of the digital twin model, including the association between the time series and spatial location of the operation data;

[0078] Based on the real-time data-driven digital twin model, the operating status of the compressor is predicted, the key state variables in the model are updated in real time, and operating feedback matching the physical system is generated.

[0079] The digital twin unit of the industrial cloud platform is the core module for building and maintaining the digital twin model of the compressor system. It has full-process functions from data reception, decryption, initial model establishment to real-time data mapping and state prediction. Its main goal is to achieve accurate modeling and real-time feedback of the operating status through synchronous operation with the physical compressor system.

[0080] The digital twin unit first receives the encrypted operating data transmitted from the IoT gateway unit. To ensure the validity and security of the data, the unit decrypts the received data and performs multiple verification operations, including integrity verification, source legitimacy check, and time synchronization check. Integrity verification uses the checksum of the data packet to ensure that the data has not been tampered with or lost during transmission; the source legitimacy check ensures that the data comes from an authorized device by checking the device identification and system registration list; the time synchronization check confirms whether the timestamp of the data matches the system clock to avoid inconsistencies caused by delayed or misordered transmission.

[0081] When the system starts or receives operating data for the first time, the digital twin unit establishes an initial digital twin model based on the physical characteristics and structural parameters of the compressor. This model covers basic information about the geometry, operating status, and dynamic parameters of the compressor. For example, the geometry may include the cavity shape, valve layout, and intake duct design of the compressor; the operating status records the initial operating parameters of the compressor, such as standard operating temperature, pressure, and vibration level; and the dynamic parameters include performance indicators that need to be updated in real time, such as the current input power, output power, and load conditions. This initial model provides a structured framework and reference point for the dynamic update of the digital twin.

[0082] As real-time data is received and processed, the digital twin unit maps the decrypted and verified operating data to the dynamic parameters of the digital twin model. This mapping process requires associating the time series data with the physical spatial location of the compressor. For example, vibration data may be associated with a specific bearing location, while temperature data may be associated with the location of the intake or exhaust port. With this precise parameter mapping, the digital twin model can reflect the dynamic changes of the physical compressor in real time.

[0083] Based on the digital twin model driven by real-time data, the digital twin unit predicts the operating status of the compressor and updates the key state variables in the model in real time. The operating status prediction uses a combination of real-time data and historical data to identify potential abnormal patterns or trends. For example, an abnormal increase in vibration amplitude may indicate mechanical wear, or a decrease in output power may indicate reduced energy efficiency. By updating these state variables, the model is able to generate operational feedback that matches the physical system, such as recommending adjustments to pressure setpoints or optimizing load distribution strategies. These feedbacks can be directly passed to downstream modules, such as intelligent diagnostic optimization units, to support further fault diagnosis and parameter optimization.

[0084] The digital twin unit accurately reflects the dynamic changes of the physical system in the digital model through steps such as decryption, verification, modeling and real-time updating, while providing operation prediction and feedback functions.

[0085] Furthermore, the industrial cloud platform digital twin unit uses stream computing technology to analyze the real-time operation data in the digital twin model, and extracting key performance indicator data includes the following steps:

[0086] Divert the real-time operation data in the digital twin model to multiple computing nodes, and perform classification processing according to the data type, including time alignment, data filtering and outlier correction; wherein the data type includes temperature, pressure, vibration and current;

[0087] Based on the distributed computing architecture, the classified data streams are aggregated in real time to generate a summary data stream representing the overall operating characteristics of the compressor;

[0088] The key performance indicator data is calculated according to the preset analysis logic for the real-time aggregated data stream, including: calculating the energy efficiency ratio based on the ratio of the compressor input power to the output power; evaluating the calculation reliability based on the comparison of historical data including running time, load conditions and vibration characteristics with current data; calculating the health degree based on the deviation degree of real-time monitored temperature, pressure and vibration;

[0089] When calculating key performance indicators, the flow computing rule base is combined to mark abnormal features in real time, including sudden vibration exceeding the limit, drastic temperature fluctuations, or pressure deviation from the normal range.

[0090] The digital twin unit of the industrial cloud platform uses stream computing technology to analyze the real-time operating data in the digital twin model, achieving accurate extraction of key performance indicators and real-time marking of abnormal features. This process relies on an efficient data processing architecture and scientific analysis logic, which can dynamically evaluate the operating status of the compressor system and provide a basis for subsequent diagnosis and optimization.

[0091] In the initial stage of analysis, the digital twin unit diverts real-time operating data from the digital twin model to multiple computing nodes. The diversion operation is implemented through the system's distributed computing architecture to ensure high data throughput and low-latency processing performance. After the data is diverted, it is classified and processed according to the data type, including time alignment, data filtering, and outlier correction operations. For example, for vibration signals, by aligning the timestamps of the data, it can ensure that data from different sources are processed synchronously; for temperature or pressure signals, the data filtering step is used to eliminate instantaneous deviations that may be caused by sensor noise; and outlier correction corrects the data by referring to the historical operating mode of the equipment to make it conform to the actual physical laws. The data types after classification include temperature, pressure, vibration, and current, which together constitute the key dynamic parameters of compressor operation.

[0092] After completing the classification process, the unit aggregates the processed data stream in real time based on the distributed computing architecture. The aggregation operation integrates the classified data from different sensors and computing nodes to generate a summary data stream that can reflect the overall operating characteristics of the compressor. Real-time aggregation ensures the high timeliness of the data while reducing analysis errors caused by data redundancy or dispersion. The aggregated data stream is passed to the analysis module in a structured form, laying the data foundation for the calculation of key performance indicators.

[0093] Next, the system calculates key performance indicators according to the preset analysis logic, including energy efficiency ratio, reliability and health. Among them, the energy efficiency ratio is dynamically calculated by the ratio of compressor input power to output power, reflecting the changes in compressor operating efficiency; reliability is obtained by analyzing the operating time, load conditions and vibration characteristics, and combining historical data with current data for comparative evaluation. It can characterize the possibility of failure of the compressor under the current operating state; health is calculated based on the deviation of real-time monitored temperature, pressure and vibration data from the normal operating range, and is mainly used to evaluate the stability and health of the overall system operation.

[0094] In the process of calculating key performance indicators, the system combines the stream computing rule base to mark abnormal features in the data in real time. The rule base contains preset conditions and thresholds for common compressor anomalies, such as exceeding the limit of vibration signals, drastic temperature fluctuations, or abnormal pressure deviations. Once the real-time data triggers these conditions, the system will immediately mark the relevant features and generate alarm information. For example, if a sensor detects that the vibration amplitude is significantly higher than the normal value, the anomaly will be marked in real time and passed to the subsequent diagnosis module together with the current performance indicators. This real-time marking mechanism can effectively intervene before potential problems expand, thereby improving the safety and reliability of the entire system.

[0095] In summary, the multi-dimensional analysis of real-time operating data in the digital twin model through stream computing technology ensures the efficient extraction of key performance indicators, and combined with the abnormal marking mechanism, realizes the dynamic evaluation of the compressor operating status.

[0096] The intelligent diagnosis and optimization unit 104 is used to receive the key performance indicators transmitted by the digital twin unit of the industrial cloud platform; based on the key performance indicators, high-order spectrum analysis is used to perform fault diagnosis on the operating status of the compressor to generate fault diagnosis results; combining the fault diagnosis results with the key performance indicators, the operating parameters of the compressor are optimized using a reinforcement learning algorithm to generate optimized operating parameters, and the optimized operating parameters include a pressure setting value, a load distribution strategy, and an energy efficiency optimization plan.

[0097] The intelligent diagnosis and optimization unit 104 is a key component. Its main function is to receive key performance indicator data transmitted by the digital twin unit of the industrial cloud platform, combine high-order spectrum analysis technology to perform fault diagnosis, and use reinforcement learning algorithm to optimize the operating parameters of the compressor, and generate optimized operating parameters to guide the dynamic adjustment of the system.

[0098] When the intelligent diagnosis and optimization unit receives the key performance indicator data transmitted by the digital twin unit of the industrial cloud platform, it will first parse and classify the data. The key performance indicator data includes energy efficiency ratio, reliability and health, which provide detailed information on the current operating status of the compressor, such as the efficiency ratio of power input and output, the stability of vibration characteristics, and whether the temperature and pressure are within the healthy range. The parsed data will be stored in the local memory for subsequent processing.

[0099] During the fault diagnosis phase, the intelligent diagnosis optimization unit uses high-order spectrum analysis technology to conduct in-depth analysis of the operating data. By analyzing the nonlinear and non-Gaussian characteristics in the vibration signal, the unit can detect potential fault features that are difficult to capture with traditional methods. For example, when certain mechanical faults (such as bearing wear or valve failure) occur, unique harmonic coupling features are generated in the vibration signal. The high-order spectrum analysis module can extract these features and compare them with the preset fault mode library to generate diagnostic results of fault type and risk level.

[0100] After completing the fault diagnosis, the intelligent diagnosis and optimization unit will combine the fault diagnosis results and key performance indicator data to optimize the operating parameters of the compressor through the reinforcement learning algorithm. The core of the reinforcement learning module is to optimize system performance through the interaction of state, action and reward. Specifically, the operating state is composed of key performance indicator data and fault diagnosis results, including the current energy efficiency level, load distribution, health status and fault risk. Based on these states, the reinforcement learning module will evaluate a variety of possible adjustment actions, such as changing the pressure set point, redistributing the load or selecting different energy efficiency optimization strategies. Each adjustment action will be evaluated according to the system's preset reward function, which mainly comprehensively considers energy efficiency improvement, fault risk reduction and system operation stability. The reinforcement learning module gradually determines the optimal operating parameters through repeated trial and error and learning.

[0101] The optimized operating parameters include pressure setting value, load distribution strategy and energy efficiency optimization scheme. The pressure setting value is the target pressure range adjusted according to different operating conditions, aiming to ensure efficient operation of the compressor and reduce energy consumption. The load distribution strategy is used to optimize the load distribution ratio of multiple compressors to achieve balanced operation and maximize energy efficiency of the system as a whole. The energy efficiency optimization scheme adjusts the power output and operating mode according to historical data and real-time operating status to further improve system efficiency.

[0102] The generated optimized operating parameters will be immediately transmitted to the distributed autonomous control unit 105 to guide the adjustment of the real-time operating status of the compressor. Through the dynamic application of these optimized parameters, the system can significantly reduce the risk of potential failures and extend the service life of the equipment while maintaining high energy efficiency.

[0103] In summary, the intelligent diagnosis and optimization unit 104 realizes comprehensive diagnosis and optimization adjustment of the compressor operating status through the combination of high-order spectrum analysis and reinforcement learning algorithm.

[0104] Furthermore, the intelligent diagnosis optimization unit uses a high-order spectrum analysis method to perform fault diagnosis on the compressor operation state, including the following steps:

[0105] Receive the operating vibration signal x(t) and key performance indicators transmitted by the digital twin unit of the industrial cloud platform, where t is the time variable;

[0106] According to the following formula (1), the bispectrum B(f1, f2) is calculated based on the operating vibration signal x(t):

[0107]

[0108] Among them, X n (f) is the result of the segmented Fourier transform of x(t); N is the total number of signal segments; (f) is X n (f) is the complex conjugate; f1 and f2 are frequency pairs used to analyze nonlinear coupling relationships;

[0109] According to the following formula (2), the normalized bispectral amplitude is calculated:

[0110]

[0111] According to the following formula (3), the dynamic significance threshold T is set B (f):

[0112] T B (f) = μ B (f)+k·σ B (f) (3)

[0113] Among them, μ B (f) is the mean amplitude around frequency f; σ B (f) is the standard deviation of the amplitude near frequency f; k is the significance factor;

[0114] From the normalized bispectral amplitude Extract the value exceeding the dynamic threshold T B The frequency of (f) fault ,f normal );

[0115] According to the normalized bispectral amplitude and the key performance indicators to generate fault diagnosis results.

[0116] The intelligent diagnosis and optimization unit in this embodiment uses a high-order spectrum analysis method to perform fault diagnosis on the operating status of the compressor, providing a technical solution for accurately identifying the nonlinear coupling characteristics and potential faults of the system. High-order spectrum analysis can capture nonlinear features that are difficult to detect with traditional frequency domain analysis, especially when the vibration signal contains harmonic coupling, noise or other complex features, this method shows unique advantages.

[0117] First, the intelligent diagnosis and optimization unit receives the operating vibration signal x(t) and key performance indicators from the digital twin unit of the industrial cloud platform. The vibration signal x(t) is a time series signal recorded by the sensor when the compressor is running, where t represents the time variable. Key performance indicators include energy efficiency ratio, health and reliability, etc., which are used to provide additional contextual information and operating background for the diagnosis results.

[0118] In order to extract the nonlinear features in the vibration signal, we first perform a piecewise Fourier transform on x(t) to obtain the frequency domain representation of each segment of the signal X n (f), where n is the segment index of the signal and f is the frequency. Then, the bispectrum B(f1,f2) is calculated according to formula (1), specifically:

[0119]

[0120] here, Yes X n The complex conjugate of (f) is used to reflect the energy distribution and phase relationship of the signal. The frequency pair (f1, f2) represents the coupling relationship between different frequencies. The bispectrum reveals nonlinear characteristics by analyzing these coupling relationships. For example, if a nonlinear vibration fault occurs in a compressor component (such as bearing failure), the fault may be manifested in the bispectrum as a significant enhancement of the coupling of a specific frequency.

[0121] Next, the normalized bispectral amplitude is calculated according to formula (2):

[0122]

[0123] In this formula, the bispectral amplitude |B(f1,f2)| is normalized to a range with a maximum value of 1 to eliminate the influence of the absolute value of the amplitude on the analysis, and the sensitivity to significant frequency pairs is enhanced through logarithmic transformation ln(1+|B(f1,f2)|). For example, if the amplitude of a frequency pair (f1,f2) is significantly higher than that of other frequency pairs, its normalized result will stand out and be easier to identify.

[0124] In order to screen out the significant frequency pairs related to faults, the dynamic significance threshold T is set according to formula (3): B (f):

[0125] T B (f) = μ B (f)+k·σ B (f)

[0126] Among them, μ B (f) is the mean value of the amplitude near frequency f, and the frequency near frequency can be frequency f±1Hz. σ B(f) is the standard deviation, and k is the significance factor, which is used to control the sensitivity of the threshold. For example, when k is set larger, the threshold is higher and only significantly abnormal frequency pairs are identified; conversely, a lower k can detect more frequency pairs, but may increase the false alarm rate.

[0127] By normalizing the bispectral amplitude With dynamic threshold T B (f) can be compared to filter out the frequency pairs that exceed the threshold (f fault ,f normal ), where f fault is the frequency associated with the potential fault, f normal It is the frequency characteristic of the system when it is operating normally.

[0128] Finally, combined with the normalized bispectral amplitude The combination of the two approaches improves the accuracy and interpretability of the diagnosis.

[0129] For example, after the vibration signal is processed, a significant frequency pair (f fault ,f normal ),in:

[0130] f fault =15Hz (potential fault related frequency).

[0131] f normal =30Hz (frequency related to normal operating characteristics).

[0132] Normalized bispectral amplitude This indicates that the coupling between the two frequencies is strong. Key performance indicators such as energy efficiency and health are also provided:

[0133] The energy efficiency ratio is 85%, slightly lower than the normal value of 90%.

[0134] The health level is 60 points (out of 100), which is significantly lower than the operating standard.

[0135] The process of generating fault diagnosis results is as follows:

[0136] 1. Analyze the normalized bispectral amplitude

[0137] It shows that there is a strong nonlinear coupling between 15Hz and 30Hz. Combined with the frequency characteristics of the vibration signal, this may be related to nonlinear faults in mechanical components, such as rotor imbalance or bearing degradation.

[0138] 2. Combine key performance indicators

[0139] The energy efficiency ratio dropped to 85%, indicating that the compressor's operating efficiency has decreased; the health score is only 60 points, indicating that the overall condition of the equipment is poor. In particular, the significant decrease in health further strengthens the potential fault association found in the vibration signal analysis.

[0140] 3. Generate fault diagnosis results

[0141] Based on the normalized bispectral amplitude and key performance indicators, the fault diagnosis module generates the following conclusions:

[0142] Fault type: It is presumed that the nonlinear vibration is enhanced due to bearing degradation.

[0143] Fault risk level: Medium (determined based on health score and bispectral amplitude strength).

[0144] Recommended Action: Bearing inspection or replacement is recommended to avoid further deterioration of the fault.

[0145] Furthermore, the state of the reinforcement learning algorithm includes a fault diagnosis result, a key performance indicator, a current operating parameter, and a current environmental condition; wherein the fault diagnosis result includes a fault type and a corresponding fault risk value, wherein the fault risk value is determined according to the key indicator;

[0146] The actions of the reinforcement learning algorithm include adjusting the pressure setpoint of each compressor, redistributing the load, and selecting different energy efficiency optimization schemes;

[0147] The reinforcement learning algorithm uses the reward function R provided by the following formula (4): t :

[0148]

[0149] Among them, Δη represents the improvement of energy efficiency ratio, which is used to reward the improvement of system operation efficiency; ΔP represents the pressure adjustment, which is the sum of the absolute values ​​of the pressure changes of all compressors in the current time step, which is used to punish excessive operation adjustments; ΔL balance L is the load balance improvement value, which is the reduction in the difference in compressor load distribution, and is used to reward a more balanced load distribution; total is the total system load; ΔD is the change in the fault risk value, which is used to punish the increase in fault risk; α1, α2, α3 are weight coefficients for adjusting the reward weight.

[0150] The reinforcement learning algorithm in this embodiment dynamically adjusts the system's operating strategy in the operation optimization of the compressor air system of the industrial cloud platform by combining fault diagnosis results, key performance indicators, current operating parameters and environmental conditions to maximize energy efficiency while reducing the risk of failure and the impact of uneven operating load. The design of the entire solution is based on the state-action-reward mechanism of reinforcement learning, and the reward function t is used to guide the system to achieve efficient, safe and balanced operation.

[0151] The state of the reinforcement learning algorithm includes four main aspects: fault diagnosis results, key performance indicators, current operating parameters and environmental conditions. The fault diagnosis results include the type of fault and the fault risk value. The fault risk value is a quantitative indicator calculated based on key performance indicators (such as health, reliability, etc.) to reflect the possibility of the current failure of the system. For example, when the vibration signal is abnormally enhanced and the health score drops significantly, the fault risk value may increase significantly. Key performance indicators include system performance data such as energy efficiency ratio and health; current operating parameters include compressor pressure setting value, load distribution, etc.; environmental conditions include external influencing factors such as temperature and humidity.

[0152] The actions of the reinforcement learning algorithm are designed as three main adjustment strategies, including adjusting the pressure setting value of each compressor, redistributing the load, and selecting different energy efficiency optimization schemes. For example, in a heavy load operation scenario, the algorithm may reduce the load by reducing the pressure setting value of some compressors, while increasing the load distribution ratio of other compressors to achieve load balancing for the overall operation.

[0153] Reward function R t It is the core of the reinforcement learning algorithm, used to evaluate the effect of each step of adjusting the strategy and guide the algorithm optimization direction. Formula (4) is defined as follows:

[0154]

[0155] The first term Δη represents the improvement in the energy efficiency ratio, which is used to reward improvements in the operating efficiency of the compressor system. The improvement in the energy efficiency ratio is calculated by comparing the energy efficiency ratio of the current time step with the energy efficiency ratio of the previous time step. For example, if the energy efficiency ratio of the previous time step is 85% and it is currently increased to 87%, then Δη = 2%. This reward drives the algorithm to prioritize actions that can significantly improve operating efficiency.

[0156] The second term ΔP is the absolute value of the sum of all compressor pressure changes in the current time step, which is used to penalize excessive operating adjustments. Excessive pressure adjustments may increase the risk of wear of mechanical components, so the penalty term of ΔP is introduced and normalized by ΔP / max(P). For example, if the total pressure change in a certain time step is 10kPa, and the maximum allowable pressure of the system is 100kPa, then this term is normalized to 0.1, and the algorithm will tend to choose the optimization solution with smaller pressure change.

[0157] The third term ΔL balance is the load balance improvement value, used to reward a more balanced load distribution.

[0158] Load balance improvement value ΔL balance It is an indicator to measure the change in the balance of compressor load distribution, and is used to quantify the reduction in the difference in compressor load distribution before and after adjustment. It is specifically defined as the difference between the maximum load difference between the compressors before adjustment and the maximum load difference between the compressors after adjustment.

[0159] The maximum load difference refers to the difference between the compressor with the highest load and the compressor with the lowest load among all the compressors in the system. This difference indicates the degree of imbalance in load distribution. The larger the difference, the more uneven the distribution; the smaller the difference, the more even the distribution.

[0160] When calculating the load balance improvement value, the system will first find the compressor with the highest load and the compressor with the lowest load based on the load distribution before adjustment, and calculate their load difference; then, the system will repeat the above steps based on the adjusted load distribution to calculate the new load difference. The difference between the two is the load balance improvement value ΔL balancc .

[0161] This value is used to evaluate whether the load balance has improved. If it has improved (i.e. the difference has decreased), then ΔL balanc is positive; if there is no improvement (i.e. the difference remains unchanged or increases), then ΔL balance is zero or negative. Through this indicator, the reinforcement learning algorithm can prioritize the optimization actions that are more conducive to load balancing.

[0162] Assume there are three compressors in the system, and initially, their loads are 70%, 50%, and 30%. By definition, you need to find the compressor with the highest load and the compressor with the lowest load to calculate the difference between them. Initially, the highest load is 70% and the lowest load is 30%, so the initial load difference is 70% minus 30%, which is 40%.

[0163] Next, suppose the system is adjusted and the loads of the three compressors are adjusted to 60%, 50% and 40% respectively. Similarly, according to the adjusted state, the highest load is 60% and the lowest load is 40%. The adjusted load difference is 60% minus 40%, which is 20%.

[0164] Finally, calculate the load balance improvement. This is the difference between the initial load difference and the adjusted load difference. The initial difference was 40%, and the adjusted difference was 20%, so the load balance improvement is 40% minus 20%, or 20%.

[0165] This improvement indicates that the load distribution of the compressors is more balanced through adjustment. The reinforcement learning algorithm will incorporate this positive improvement into the reward function, encouraging the system to further optimize the load distribution strategy.

[0166] The fourth term ΔD is the change in the fault risk value, which is used to penalize the increase in fault risk. For example, if the fault risk value increases from 0.2 to 0.3, then ΔD = 0.1. In order to emphasize the potential threat of fault risk to system operation, the logarithmic transformation ln(1+ΔD 2 ) further amplifies the impact of the risk value and ensures that the algorithm prioritizes actions that reduce the risk of failure.

[0167] The weight coefficients α1, α2, and α3 are used to adjust the priorities of different objectives in the reward function. For example, in the early stages of system operation, it may be necessary to pay more attention to improving energy efficiency and set a larger α1; in high-risk scenarios, it may be necessary to increase the weight of α3 to prioritize reducing the risk of failure.

[0168] In summary, the reinforcement learning algorithm can dynamically optimize the operating parameters of the compressor by combining state, action and reward functions, balancing the needs of energy efficiency improvement, load balancing and fault risk control.

[0169] In summary, this embodiment uses a high-order spectrum analysis method to perform in-depth analysis on the vibration signal of the compressor and accurately identify significant frequency characteristics related to potential faults.

[0170] The distributed autonomous control unit 105 is used to receive the optimized operating parameters and fault diagnosis results fed back by the intelligent diagnosis and optimization unit; adjust the real-time operating state of the compressor according to the optimized operating parameters to achieve load adaptive balancing and energy efficiency optimization; in the event of communication interruption, start the local autonomous control strategy according to the fault diagnosis results and execute emergency control under emergency conditions.

[0171] The distributed autonomous control unit 105 is one of the important modules, which aims to achieve dynamic adjustment of the compressor system, load balancing, and reliable operation in the event of communication interruption by executing optimized operating parameters and emergency control strategies. The unit dynamically controls the real-time operating status of the compressor by receiving optimized operating parameters and fault diagnosis results from the intelligent diagnosis and optimization unit, and performs local autonomous functions in special circumstances.

[0172] The distributed autonomous control unit first receives the optimized operating parameters and fault diagnosis results transmitted by the intelligent diagnosis and optimization unit through its communication interface. The optimized operating parameters include pressure setting values, load distribution strategies, and energy efficiency optimization schemes, which provide precise guidance for the specific operation adjustments of the compressor. The fault diagnosis results provide detailed information about the current system operation status, including possible abnormal conditions and their severity. After receiving this data, the control unit parses and distributes it to ensure that each piece of information can be quickly matched to a specific control task.

[0173] Under normal communication conditions, the distributed autonomous control unit adjusts the real-time operating status of the compressor according to the optimized operating parameters. This adjustment process relies on the load adaptive balancing logic, which dynamically allocates the load ratio of each device in the compressor group to achieve efficient operation of the entire system. For example, when a compressor is running close to its optimal efficiency point, the control unit will give priority to allocating more load to the device, while reducing the operating load of other devices to improve the energy efficiency performance of the entire system. This dynamic balancing mechanism can effectively reduce energy loss in the system, while avoiding overload operation of a single compressor and increasing the service life of the equipment.

[0174] While processing and optimizing operating parameters, the distributed autonomous control unit also monitors the operating status of the compressor in real time to ensure that the system's adjustment actions will not cause new operating anomalies. For example, when the load distribution strategy is executed, the control unit will immediately check whether the operating parameters of all compressors are still within the normal range, such as temperature, pressure, vibration and other indicators. Once it is found that some parameters are out of the safe range, the system will issue a warning and readjust the load distribution strategy to prevent the fault from further deteriorating.

[0175] When a communication interruption occurs, the distributed autonomous control unit initiates a local autonomous control strategy to ensure the continued operation of the system. At this point, the control unit relies on the historical operating data and optimized parameters stored locally, combined with the fault diagnosis results, to independently complete the operation control of the compressor. For example, if the diagnosis results indicate that a certain device has a high-risk fault, the control unit will prioritize reducing the operating load of the device or shut it down directly, while distributing the load to other compressors that are operating normally. In addition, for devices where no fault risk is detected, the control unit will maintain its operation according to the optimal parameters in the historical records to ensure that the basic functions of the system are maintained.

[0176] In order to deal with emergencies, the distributed autonomous control unit also has emergency response capabilities. For example, when a compressor is detected to be vibrating violently or the temperature is rising sharply, the control unit can immediately take protective measures, such as quickly reducing the compressor speed or cutting off the power supply, to avoid equipment damage or safety accidents. This emergency strategy can run independently when the system is out of the support of the cloud platform, greatly improving the reliability of the system.

[0177] In summary, the distributed autonomous control unit enhances the stability of the system under special circumstances while ensuring efficient operation of the system through dynamic load balancing and intelligent local control strategies.

[0178] Furthermore, the distributed autonomous control unit realizes load adaptive balancing including:

[0179] Receive and analyze the optimized operating parameters fed back by the intelligent diagnosis and optimization unit, including the pressure setting value and current load of each compressor;

[0180] Based on the real-time monitoring of the load distribution status, dynamically adjust the load distribution ratio of each compressor in the compressor group to ensure balanced distribution while keeping the total load unchanged;

[0181] Prioritize the most efficient compressors during load distribution and reduce the load of less efficient compressors to achieve overall energy efficiency optimization;

[0182] The adjusted load distribution status is continuously monitored, and adaptive balancing calculations are re-executed when load imbalance or parameter abnormalities occur.

[0183] The distributed autonomous control unit in this embodiment realizes the load adaptive balancing function, so that the industrial compressor air system can dynamically adjust the load distribution ratio according to the real-time status, thereby optimizing the overall operating efficiency. Its core lies in receiving optimization parameters, real-time monitoring of system status, dynamic adjustment of load distribution, and continuous feedback optimization. These steps constitute a closed-loop control mechanism to ensure that the system is always in an efficient and stable state during operation.

[0184] The distributed autonomous control unit first receives the optimized operating parameters fed back by the intelligent diagnosis and optimization unit. These parameters include the pressure setting value and current load of each compressor. The pressure setting value reflects the target operating state of each compressor under the current optimization strategy, while the current load describes the workload distribution of the compressor in actual operation. By comprehensively analyzing this information, the distributed autonomous control unit can fully understand the current state of the system and provide an accurate reference for subsequent load adjustments.

[0185] During the load adjustment process, the distributed autonomous control unit will dynamically adjust the load ratio of each compressor in the compressor group based on the real-time monitored load distribution status. The principle of adjustment is to balance the workload of each compressor as much as possible under the premise that the total load remains unchanged. For example, in a system with three compressors, if the load of one compressor is much higher than that of other compressors, the system will reduce the load of the compressor and redistribute the excess load to other compressors to achieve the purpose of overall load uniformity. This dynamic adjustment can effectively reduce the risk of overload of a single compressor and extend the service life of the equipment.

[0186] When performing load distribution, the distributed autonomous control unit also gives priority to the compressors with the highest operating efficiency. Specifically, the system analyzes the operating efficiency of each compressor and allocates more load to efficient equipment while reducing the load on less efficient equipment. This strategy not only reduces the overall energy consumption of the system, but also ensures that the equipment operates in the best working condition. For example, if a compressor is more efficient than other equipment, the system may increase its load appropriately and reduce the operating time of inefficient equipment, thereby optimizing overall energy efficiency.

[0187] In addition, the distributed autonomous control unit has the ability to continuously monitor. After the adjustment is completed, the system will monitor the load distribution status of each compressor in real time to ensure that the adjusted distribution ratio meets expectations in actual operation. If the load is found to be unbalanced again, or if some operating parameters are abnormal due to environmental changes, such as the pressure setting value cannot meet the operating requirements, the system will immediately trigger the re-execution of the adaptive balance calculation. This closed-loop mechanism of continuous monitoring and adjustment ensures the efficient and stable operation of the system in a dynamic environment.

[0188] In summary, the distributed autonomous control unit achieves adaptive load balancing of the compressor unit by receiving optimization parameters, dynamically adjusting the load ratio, prioritizing the load of high-efficiency equipment, and continuously monitoring the adjustment effect. Its design can significantly improve the energy efficiency of the system, reduce equipment wear, and meet the requirements of industrial environments for high reliability and high efficiency.

[0189] Furthermore, the local autonomous control strategy initiated by the distributed autonomous control unit in the case of communication interruption includes the following steps:

[0190] Based on the fault diagnosis results and equipment operation history data, first determine whether there is a serious fault in the current state;

[0191] If a serious fault exists, immediately reduce the operating load to a safe range and stop the affected compressor;

[0192] If there is no serious fault, the operating state of the compressor is adjusted based on the locally stored optimized operating parameters to maintain the basic functions of the system;

[0193] The operating status data during the communication interruption is recorded and synchronized to the industrial cloud platform after the communication is restored for subsequent analysis and supplementary optimization.

[0194] The distributed autonomous control unit of this embodiment starts the local autonomous control strategy in the case of communication interruption to ensure that the industrial compressor air system can operate safely and reliably under the condition of network disconnection and maintain the basic functions of the system to the maximum extent. The core is to dynamically adjust the operating status by combining the fault diagnosis results, equipment historical data and locally stored optimization parameters, and record the system's operating data to provide support for subsequent analysis and optimization.

[0195] In the initial stage of communication interruption, the distributed autonomous control unit first prioritizes whether there is a serious fault in the current system state based on the fault diagnosis results and the equipment's operating history data. The fault diagnosis results provide information about the health of the equipment, including the specific fault type and fault risk value, while the equipment's operating history data reflects its long-term performance change trend and the possibility of faults. By comprehensively analyzing this data, the system can quickly identify potential serious problems, such as mechanical wear of key components, excessive temperature or abnormal vibration. This priority judgment process provides a basis for the selection of subsequent system operations.

[0196] If the judgment result shows that there is a serious fault, the system will immediately take protective measures to reduce the operating load to a safe range to ensure that the core components of the equipment will not be further damaged. For compressors affected by the fault, the system will directly shut down to avoid expanding the scope of the fault. For example, when the vibration signal of a compressor exceeds the allowable range and historical data shows that its bearings are at risk of degradation, the system will prioritize stopping the operation of the equipment and reduce the overall system pressure by reducing the load of other equipment. This process is designed to control potential risks to the minimum range and ensure the safety of equipment and personnel.

[0197] If the judgment result shows that there is no serious fault, the system will adjust the operating status of the compressor based on the locally stored optimized operating parameters. These optimization parameters are generated and stored by the intelligent diagnosis and optimization unit when communication is normal, including information such as pressure set points, load distribution strategies, and energy efficiency optimization plans. By referring to these parameters, the system can continue to execute the optimized operating strategy during communication interruptions and maintain the basic functions of the system. For example, under the premise of ensuring the total load demand of the system, the system may give priority to allocating loads to compressors with higher efficiency, while reducing the operating pressure of less efficient equipment, thereby maintaining efficient operation under conditions of limited resources.

[0198] During the communication interruption, the distributed autonomous control unit will continuously record the system's operating status data, including the operating load, pressure setting value, fault diagnosis results, and energy efficiency performance of each compressor. These data can not only provide a complete operating record after communication is restored, but can also be used for subsequent analysis and supplementary optimization. For example, by analyzing the operating data during the communication interruption, the system can identify potential inefficiencies or uneven load distribution problems, providing a reference for further improving the operating strategy.

[0199] After communication is restored, the distributed autonomous control unit will synchronize the recorded operating status data to the industrial cloud platform for comprehensive system analysis and optimization strategy updates. This data synchronization process ensures that the cloud platform has a complete understanding of the system operation during the communication interruption, so that the operating status of the entire system can be optimized globally.

[0200] In summary, the local autonomous control strategy achieves safe system operation and maintenance of basic functions in the event of communication interruption by combining fault diagnosis results, locally stored optimization parameters and operation history data.

[0201] Although the present application is disclosed as above in the form of a preferred embodiment, it is not intended to limit the present application. Any technical personnel in this field may make possible changes and modifications without departing from the spirit and scope of the present application. Therefore, the scope of protection of the present application shall be based on the scope defined by the claims of the present application.

Claims

1. A control system for compressor air system of industrial cloud platform based on Internet of Things gateway, characterized in that: include: A data acquisition unit is used to collect the operating data of the compressor in real time through intelligent sensors configured at the air inlet end, the exhaust end and the body of the compressor, wherein the operating data includes temperature, pressure, vibration and current; pre-process the collected operating data, and transmit the pre-processed operating data to the Internet of Things gateway unit; An Internet of Things gateway unit, used for receiving the pre-processed operation data transmitted by the data acquisition unit; Perform end-to-end encryption on the received data; verify the device identity of the data source and perform dynamic access control to ensure the security of data transmission; transmit the encrypted and verified operation data to the digital twin unit of the industrial cloud platform; The digital twin unit of the industrial cloud platform is used to receive the encrypted operation data transmitted by the IoT gateway unit; after decrypting and verifying the operation data, the digital twin model of the compressor system is constructed and updated in real time based on the received data, and the digital twin model is used to synchronously map the physical system and information system of the compressor; the real-time operation data in the digital twin model is analyzed by using stream computing technology to extract key performance indicator data, and the key performance indicator data includes energy efficiency ratio, reliability and health; the extracted key performance indicator data is synchronously stored with the original operation data in the time series database to support historical trend analysis, fast retrieval and data modeling optimization; Deliver the extracted key performance indicator data to the intelligent diagnosis and optimization unit; An intelligent diagnosis and optimization unit is used to receive key performance indicators transmitted by the digital twin unit of the industrial cloud platform; based on the key performance indicators, high-order spectrum analysis is used to perform fault diagnosis on the operating status of the compressor and generate fault diagnosis results; Combining the fault diagnosis results with the key performance indicators, the operating parameters of the compressor are optimized using a reinforcement learning algorithm to generate optimized operating parameters, which include a pressure setting value, a load distribution strategy, and an energy efficiency optimization solution; The distributed autonomous control unit is used to receive the optimized operating parameters and fault diagnosis results fed back by the intelligent diagnosis and optimization unit; adjust the real-time operating status of the compressor according to the optimized operating parameters to achieve load adaptive balancing and energy efficiency optimization; in the event of communication interruption, start the local autonomous control strategy according to the fault diagnosis results and execute emergency control under emergency conditions.

2. The control system of the industrial cloud platform compressor air system based on the Internet of Things gateway according to claim 1 is characterized in that: The data acquisition unit is specifically used for: Aggregate the collected data into local data cache through sensor nodes to form small batch data blocks; After the data block is formed, the data block is sent to the IoT gateway unit at fixed time intervals using a timing transmission protocol; During the transmission process, data verification information is added, including acquisition timestamp, sensor identification and data summary verification value, to ensure the integrity and correctness of the data during transmission.

3. The control system of the industrial cloud platform compressor air system based on the Internet of Things gateway according to claim 1 is characterized in that: The data acquisition unit includes a data preprocessing module, and the data preprocessing module is used to: Data denoising to remove environmental interference signals, including outlier detection based on sensor signal consistency; Data integrity check: by verifying the continuity and validity of the collected signal, the abnormal data segments that do not meet the requirements are eliminated; Data standardization converts the operating data into a uniform numerical range for subsequent processing and analysis.

4. The control system of the industrial cloud platform compressor air system based on the Internet of Things gateway according to claim 1 is characterized in that: The Internet of Things gateway unit includes the following modules: A data receiving module is used to receive the pre-processed operation data from the data acquisition unit and perform consistency check on the timestamp and source device identification of the data; An encryption module, used to perform end-to-end encryption processing on the verified data, wherein the encryption process dynamically generates a key and synchronizes the key with the digital twin unit of the industrial cloud platform; Authentication module, used to verify the identity of the data source device, including device identification comparison and historical trust record check; The dynamic access control module is used to assign or adjust access rights based on the verification results of the device identity and monitor abnormal behavior during data interaction in real time.

5. The control system of the industrial cloud platform compressor air system based on the Internet of Things gateway according to claim 1 is characterized in that: The industrial cloud platform digital twin unit constructs and updates the digital twin model of the compressor system in real time based on the received data, including the following steps: After receiving the encrypted operation data transmitted from the IoT gateway unit, decrypt the data and verify the integrity, source legitimacy and time synchronization of the data; When the system is started or data is first received, an initial digital twin model is established based on the physical characteristics and structural parameters of the compressor, the model including basic information on geometric structure, operating status and dynamic parameters; Mapping the decrypted and verified real-time operation data to the dynamic parameters of the digital twin model, including the association between the time series and spatial location of the operation data; Based on the real-time data-driven digital twin model, the operating status of the compressor is predicted, the key state variables in the model are updated in real time, and operating feedback matching the physical system is generated.

6. The control system of the industrial cloud platform compressor air system based on the Internet of Things gateway according to claim 1 is characterized in that: The industrial cloud platform digital twin unit uses stream computing technology to analyze the real-time operation data in the digital twin model, and extracting key performance indicator data includes the following steps: Divert the real-time operation data in the digital twin model to multiple computing nodes, and perform classification processing according to the data type, including time alignment, data filtering and outlier correction; wherein the data type includes temperature, pressure, vibration and current; Based on the distributed computing architecture, the classified data streams are aggregated in real time to generate a summary data stream representing the overall operating characteristics of the compressor; The key performance indicator data is calculated according to the preset analysis logic for the real-time aggregated data stream, including: calculating the energy efficiency ratio based on the ratio of the compressor input power to the output power; evaluating the calculation reliability based on the comparison of historical data including running time, load conditions and vibration characteristics with current data; calculating the health degree based on the deviation degree of real-time monitored temperature, pressure and vibration; When calculating key performance indicators, the flow computing rule base is combined to mark abnormal features in real time, including sudden vibration exceeding the limit, drastic temperature fluctuations, or pressure deviation from the normal range.

7. The control system of the industrial cloud platform compressor air system based on the Internet of Things gateway according to claim 1 is characterized in that: The intelligent diagnosis optimization unit uses a high-order spectrum analysis method to perform fault diagnosis on the compressor operation state, including the following steps: Receive the operating vibration signal x(t) and key performance indicators transmitted by the digital twin unit of the industrial cloud platform, where t is the time variable; According to the following formula (1), the bispectrum B(f1, f2) is calculated based on the operating vibration signal x(t): Among them, X n (f) is the result of the segmented Fourier transform of x(t); N is the total number of signal segments; For X n (f) is the complex conjugate; f1 and f2 are frequency pairs used to analyze nonlinear coupling relationships; According to the following formula (2), the normalized bispectral amplitude is calculated: According to the following formula (3), the dynamic significance threshold T is set B (f): T B (f)=μ B (f)+k·s B (f) (3) Among them, μ B (f) is the mean amplitude around frequency f; σ B (f) is the standard deviation of the amplitude near frequency f; k is the significance factor; From the normalized bispectral amplitude Extract the value exceeding the dynamic threshold T B The frequency of (f) fault ,f normal ); According to the normalized bispectral amplitude and the key performance indicators to generate fault diagnosis results.

8. The control system of the compressor air system of the industrial cloud platform based on the Internet of Things gateway according to claim 1 is characterized in that: The state of the reinforcement learning algorithm includes a fault diagnosis result, a key performance indicator, a current operating parameter, and a current environmental condition; wherein the fault diagnosis result includes a fault type and a corresponding fault risk value, wherein the fault risk value is determined according to the key indicator; The actions of the reinforcement learning algorithm include adjusting the pressure setpoint of each compressor, redistributing the load, and selecting different energy efficiency optimization schemes; The reinforcement learning algorithm uses the reward function R provided by the following formula (4): t : Among them, Δη represents the improvement of energy efficiency ratio, which is used to reward the improvement of system operation efficiency; ΔP represents the pressure adjustment, which is the sum of the absolute values ​​of the pressure changes of all compressors in the current time step, which is used to punish excessive operation adjustments; ΔL balance L is the load balance improvement value, which is the reduction in the difference in compressor load distribution, and is used to reward a more balanced load distribution; total is the total system load; ΔD is the change in the fault risk value, which is used to punish the increase in fault risk; α1, α2, α3 are weight coefficients for adjusting the reward weight.

9. The control system of the compressor air system of the industrial cloud platform based on the Internet of Things gateway according to claim 1 is characterized in that: The distributed autonomous control unit realizes load adaptive balancing including: Receive and analyze the optimized operating parameters fed back by the intelligent diagnosis and optimization unit, including the pressure setting value and current load of each compressor; Based on the real-time monitoring of the load distribution status, dynamically adjust the load distribution ratio of each compressor in the compressor group to ensure balanced distribution while keeping the total load unchanged; Prioritize the most efficient compressors during load distribution and reduce the load of less efficient compressors to achieve overall energy efficiency optimization; The adjusted load distribution status is continuously monitored, and adaptive balancing calculations are re-executed when load imbalance or parameter abnormalities occur.

10. The control system of the compressor air system of the industrial cloud platform based on the Internet of Things gateway according to claim 1 is characterized in that: The local autonomous control strategy initiated by the distributed autonomous control unit in the case of communication interruption includes the following steps: Based on the fault diagnosis results and equipment operation history data, first determine whether there is a serious fault in the current state; If a serious fault exists, immediately reduce the operating load to a safe range and stop the affected compressor; If there is no serious fault, the operating state of the compressor is adjusted based on the locally stored optimized operating parameters to maintain the basic functions of the system; The operating status data during the communication interruption is recorded and synchronized to the industrial cloud platform after the communication is restored for subsequent analysis and supplementary optimization.

Images