Instant messaging method of cloud communication platform

Abstract

The invention provides an instant messaging method for a cloud communication platform, and the method comprises the steps: obtaining the equipment type, production technology and operation parameters of industrial equipment, judging whether the equipment has a fault symptom or not in combination with a pre-established equipment fault mode library, and determining the fault type and the fault severity if the fault symptom exists; and acquiring a network state and an equipment load condition, calculating an optimal data transmission path and a server node in real time through a pre-established load balancing method, and determining the matched data transmission path and server node from the path and node mapping table according to the acquired data transmission real-time performance requirement and data transmission quality requirement.

Description

Technical Field

[0001] The present invention relates to the field of information technology, and in particular to an instant messaging method for a cloud communication platform. Background Art

[0002] In the process of remote diagnosis and maintenance of industrial equipment, it is necessary to transmit equipment operation data, fault information, and audio and video data through instant messaging platforms. However, equipment failures of different types and severity have different requirements for the real-time and quality of data transmission. For example, for serious equipment failures, high-quality, low-latency audio and video transmission is required for experts to conduct remote diagnosis and guidance, and portable diagnostic instruments such as vibration analyzers and thermal imagers need to be quickly connected to obtain detailed status data of the equipment. For general equipment failures, the requirements for audio and video quality and real-time performance of diagnostic instrument data are relatively low. In addition, industrial control systems and manufacturing execution systems have extremely high requirements for network security and stability, and any improvement in communication quality and efficiency cannot be at the expense of system security. Therefore, how to adjust the priority and quality of data transmission in real time according to the severity of equipment failures and diagnostic requirements, balance network load to maximize communication efficiency, and flexibly allocate and recycle network resources between different diagnostic tasks under the premise of ensuring the security of industrial control systems and manufacturing execution systems is a technical problem faced by the field of remote diagnosis and maintenance of industrial equipment. Summary of the invention

[0003] The present invention provides an instant messaging method for a cloud communication platform, which mainly includes:

[0004] Obtain the equipment type, production process and operating parameters of industrial equipment, and combine with the pre-established equipment failure mode library to determine whether the equipment has any signs of failure. If there are signs of failure, determine the type of failure and the severity of the failure;

[0005] According to the determined fault type and fault severity, the corresponding data transmission real-time requirements and data transmission quality requirements are obtained from the preset data transmission requirement mapping table. When the fault type is damage to a key component and the fault severity reaches a high-risk threshold, the video transmission resolution and frame rate are increased. When the fault type is general wear and tear and the fault severity does not reach a medium-risk threshold, the video transmission resolution and frame rate are reduced.

[0006] According to the determined fault type and fault severity, as well as the equipment type and production process, the diagnostic instrument is determined from the preset diagnostic mapping table. When the fault type involves rotating machinery and the fault severity reaches the high-risk threshold, a diagnostic channel is established with the vibration diagnostic instrument through the instant messaging platform. When the fault type involves static equipment and the fault severity does not reach the medium-risk threshold, an infrared thermal imager is used for diagnosis, and the conventional data transmission channel is maintained;

[0007] Obtain the diagnostic data types and data features collected by different diagnostic instruments. For vibration data collected by portable vibration diagnostic instruments, increase the transmission priority and obtain the corresponding transmission bandwidth allocation ratio from the preset vibration data transmission configuration table. For thermal imaging data collected by infrared thermal imagers, reduce the transmission priority and obtain the corresponding transmission bandwidth allocation ratio from the preset thermal imaging data transmission configuration table.

[0008] Obtain network status and device load conditions, calculate the optimal data transmission path and server node in real time through a pre-established load balancing method, and determine the matching data transmission path and server node from the path and node mapping table based on the acquired data transmission real-time requirements and data transmission quality requirements;

[0009] Based on the data security requirements of industrial control systems and manufacturing execution systems, a multi-level security protection system is built to encrypt data transmitted through instant messaging using the national secret SM4 encryption. The data access rights corresponding to diagnostic personnel at different levels are obtained from the preset diagnostic personnel authority mapping table. When diagnostic personnel are monitored to access sensitive production parameters or operate key actuators, blocking measures are taken to allow only authorized diagnostic personnel to access diagnostic data of the corresponding level and issue control instructions.

[0010] The technical solution provided by the embodiment of the present invention may have the following beneficial effects:

[0011] The present invention discloses a method for instant communication of a cloud communication platform. The method obtains device information and determines fault signs in combination with a fault mode library, and dynamically adjusts data transmission requirements according to the fault type and severity. In response to different fault conditions, the present invention flexibly selects diagnostic instruments and establishes corresponding data transmission channels, such as using a vibration diagnostic instrument for high-risk faults of rotating machinery and increasing the transmission priority. At the same time, the present invention considers the network status and equipment load, and calculates the optimal transmission path and server node in real time. In terms of data security, the present invention constructs a multi-level protection system, encrypts the transmitted data, and strictly controls the data access rights of diagnostic personnel. This intelligent fault diagnosis and data transmission method can flexibly adjust the diagnostic strategy and transmission method according to the fault situation, and improves the efficiency and accuracy of fault diagnosis while ensuring data security. BRIEF DESCRIPTION OF THE DRAWINGS

[0012] Figure 1 The present invention is a flowchart of a method for instant messaging on a cloud communication platform.

[0013] Figure 2 A schematic diagram of an instant messaging method for a cloud communication platform according to the present invention.

[0014] Figure 3This is another schematic diagram of an instant messaging method for a cloud communication platform according to the present invention. DETAILED DESCRIPTION

[0015] In order to make the purpose, technical solutions and advantages of the present invention more clear, the present invention is described in detail below with reference to the accompanying drawings and specific embodiments.

[0016] like Figure 1-3 In this embodiment, a method for instant messaging on a cloud communication platform may specifically include:

[0017] S101. Obtain the equipment type, production process and operation parameters of the industrial equipment, and determine whether the equipment has any signs of failure in combination with a pre-established equipment failure mode library. If there are signs of failure, determine the type of failure and the severity of the failure.

[0018] The equipment type, production process and operation parameter data of industrial equipment are obtained. The operation parameter data includes the real-time status information of the equipment. The collected raw data is preprocessed according to the operation parameter data, including removing outliers, filling missing values ​​and data standardization, to obtain a standardized equipment operation data set. For the standardized equipment operation data set, the statistical characteristics of each parameter, including mean, standard deviation, skewness and kurtosis, are calculated by time series analysis method, and compared with the pre-set normal parameter threshold to determine whether the equipment is in an abnormal state. If the equipment is judged to be in an abnormal state, the abnormal state is compared with the preset fault symptom library to determine whether it is a potential fault. If it is determined to be a potential fault, the corresponding fault feature template is extracted from the pre-established equipment fault mode library according to the equipment type and production process. The decision tree algorithm is used to match the potential fault with the fault feature template to determine the specific fault type of the equipment. For the specific fault type, the fuzzy logic controller is used to combine the deviation degree and duration of the equipment operation parameters to quantitatively evaluate the severity of the fault, which is divided into three levels: slight, medium and severe, and the fault diagnosis result report is output.

[0019] Exemplarily, the equipment type, production process and operation parameter data of industrial equipment are obtained from the equipment control system through the OPCUA protocol, and the collected raw data are preprocessed, including removing outliers, filling missing values ​​and data standardization, to obtain a standardized equipment operation data set. For the standardized equipment operation data set, a time series analysis method such as moving average or exponential smoothing is used to calculate the statistical characteristics of each parameter, including mean, standard deviation, skewness and kurtosis, and compared with the pre-set normal parameter threshold to determine whether the equipment is in an abnormal state. The detected anomaly is compared with the preset fault symptom library to determine whether it is a potential fault. If it is determined to be a potential fault, the corresponding fault feature template is extracted from the pre-established equipment fault mode library according to the equipment type and production process. The determined potential fault is matched with the fault feature template using a decision tree algorithm to determine the specific fault type of the equipment. For the determined fault type, a fuzzy logic controller is used to combine the deviation degree and duration of the equipment operation parameters to quantitatively evaluate the severity of the fault, which is divided into three levels: slight, medium and severe, and a fault diagnosis result report is output. The equipment type of industrial equipment, such as centrifugal pumps, production processes, such as sewage treatment, and operating parameter data, such as speed, flow, and pressure, are obtained from the Siemens S7-1500PLC control system through the OPCUA protocol, with a sampling frequency of 1Hz. The collected raw data are preprocessed, and the 3σ principle is used to remove outliers, and the linear interpolation method is used to fill missing values. Then, the data is standardized to the [-1,1] interval using the Z-score standardization method to obtain a standardized equipment operation data set. For the standardized equipment operation data set, the moving average of a 5-minute window and the exponential smoothing method with α=0.2 are used to calculate the statistical characteristics of each parameter, including mean, standard deviation, skewness, and kurtosis. The calculation results are compared with the pre-set normal parameter thresholds, such as the speed mean of 3000±100rpm, the flow standard deviation <5m 3 / h is compared to determine whether the equipment is in an abnormal state. The detected anomaly is compared with the preset fault symptom library. For example, signs such as increased speed fluctuation and decreased flow rate are associated with impeller wear faults to determine whether it is a potential fault. If it is determined to be a potential fault, the corresponding fault feature template is extracted from the pre-established equipment fault mode library according to the equipment type (centrifugal pump) and production process (wastewater treatment). For example, the impeller wear template contains features such as speed fluctuation, flow rate decrease, and efficiency reduction. A decision tree algorithm, such as the C4.5 algorithm, is used to match the determined potential fault with the fault feature template. The nodes of the decision tree contain attributes such as speed fluctuation amplitude, flow rate decrease percentage, and efficiency reduction degree. The specific fault type of the equipment is determined by traversing the decision tree. For the determined fault type, a fuzzy logic controller is used to quantitatively evaluate the severity of the fault in combination with the degree of deviation and duration of the equipment operating parameters. For example, for impeller wear fault, the fuzzy set is set to {mild wear, moderate wear, severe wear}, and the input variables are the flow rate drop percentage and duration. The membership of the fault severity is obtained through fuzzy reasoning rules, such as IF the flow rate drops by 10% and the duration is >24 hours THEN moderate wear. Finally, the centroid method is used for defuzzification, and the fault severity is divided into three levels: mild, moderate and severe. The fault diagnosis result report containing the fault type, severity and related parameters is output.

[0020] S102. According to the determined fault type and fault severity, obtain the corresponding data transmission real-time requirements and data transmission quality requirements from the preset data transmission requirement mapping table. When the fault type is damage to a key component and the fault severity reaches a high-risk threshold, increase the video transmission resolution and the frame rate. When the fault type is general wear and tear and the fault severity does not reach a medium-risk threshold, reduce the video transmission resolution and the frame rate.

[0021] The fault type and fault severity in the fault diagnosis result are obtained from a pre-established multi-dimensional array data transmission requirement mapping table, and the multi-dimensional array data transmission requirement mapping table includes data transmission real-time requirements and data transmission quality requirements; according to the fault type and fault severity, it is determined whether the key component is damaged and reaches the preset high-risk threshold; if the judgment result is that the key component is damaged and reaches the preset high-risk threshold, the video transmission resolution is increased to a preset interval and the frame rate is increased to a preset interval frame through the encoder; if the judgment result is general wear and does not reach the preset medium-risk threshold, the dynamic bit rate control function of the encoder is used to reduce the video transmission resolution to a preset interval and the frame rate to a preset interval; the video encoding parameters are dynamically adjusted using a dynamic adaptive streaming media transmission protocol; according to the real-time network status monitoring results, a version suitable for the current network conditions is selected from multiple preset bit rate versions, and the multiple bit rate versions are generated by an adaptive bit rate algorithm.

[0022] Exemplarily, according to the fault type and fault severity in the fault diagnosis result, the corresponding data transmission real-time requirements and data transmission quality requirements are queried from the pre-established multi-dimensional array data transmission demand mapping table to obtain a specific transmission parameter value range. If the fault type is judged to be damage to a key component and the fault severity reaches a preset high-risk threshold, the video transmission resolution is increased to between 1080p and 4K through the H.264 encoder, and the frame rate is increased to a range of 25 to 60 frames per second to obtain a clearer device status image. If the fault type is judged to be general wear and tear, and the fault severity does not reach the preset medium-risk threshold, the dynamic bit rate control function of the H.265 encoder is used to reduce the video transmission resolution to between 480p and 720p, and the frame rate is reduced to a range of 10 to 20 frames per second to reduce network bandwidth occupancy. According to different fault types and severities, the dynamic adaptive streaming protocol (DASH) is used to dynamically adjust video encoding parameters, including I frame interval, quantization parameter and motion estimation search range, to achieve a balance between video quality and network bandwidth. According to the real-time network status monitoring results, an adaptive bit rate algorithm is used to select the version that best suits the current network conditions from the preset multiple bit rate versions to ensure the best video quality and transmission effect in different network environments. According to the fault type in the fault diagnosis results, such as abnormal bearing wear and fault severity, such as 70%, the corresponding data transmission real-time requirements, such as delay <100ms and data transmission quality requirements, such as packet loss rate <1%, are queried from the pre-established multi-dimensional array data transmission demand mapping table to obtain the specific transmission parameter value range. If the fault type is judged to be damage to a key component, such as a broken blade, and the fault severity reaches a preset high-risk threshold, such as 85%, the video transmission resolution is increased to 1920x1080 (1080p) through the H.264 encoder, and the frame rate is increased to 30 frames per second to obtain a clearer image of the device status. If the fault type is judged to be general wear, such as aging of the seal ring, and the fault severity does not reach the preset medium-risk threshold, such as 40%, the dynamic bitrate control function of the H.265 encoder is used to reduce the video transmission resolution to 1280x720 (720p) and the frame rate to 15 frames per second to reduce network bandwidth usage. According to different fault types and severities, the dynamic adaptive streaming media transport protocol (DASH) is used to dynamically adjust the video encoding parameters, including setting the I frame interval to 2 seconds, the quantization parameter to 22, and the motion estimation search range to 16 pixels to achieve a balance between video quality and network bandwidth. According to the real-time network status monitoring results, if the current bandwidth is 5Mbps, the adaptive bitrate algorithm is used to select the 4Mbps version from the preset multiple bitrate versions of 2Mbps, 4Mbps, 6Mbps, and 8Mbps to ensure that the best video quality and transmission effect can be obtained in different network environments.

[0023] S103. According to the determined fault type and fault severity as well as the equipment type and production process, a diagnostic instrument is determined from a preset diagnostic mapping table. When the fault type involves rotating machinery and the fault severity reaches a high-risk threshold, a diagnostic channel is established with the vibration diagnostic instrument through an instant messaging platform. When the fault type involves static equipment and the fault severity does not reach a medium-risk threshold, an infrared thermal imager is used for diagnosis, and a regular data transmission channel is maintained.

[0024] According to the fault type, fault severity, equipment type and production process information contained in the fault diagnosis results, the corresponding diagnostic instrument type and diagnostic parameter configuration are obtained from the pre-established decision tree diagnostic mapping table to obtain a specific diagnostic plan; if the fault type is judged to involve rotating machinery and the fault severity reaches the preset high-risk threshold, a diagnostic instruction is sent to the on-site portable vibration diagnostic instrument, a diagnostic channel is established, and the sampling frequency and sampling duration of the vibration diagnostic instrument are set; if the fault type is judged to involve static equipment and the fault severity does not reach the preset medium-risk threshold, the infrared thermal imager is started for diagnosis, and the infrared thermal imager parameters are automatically adjusted according to the static equipment type; according to the diagnostic instrument and fault type, the collected diagnostic data is compressed and the compressed data is transmitted back to the central control system through the data transmission channel; based on the compressed diagnostic data, a decision tree algorithm is used to generate a diagnostic report containing the fault type, severity and recommended measures, and the report is sent to relevant maintenance personnel.

[0025] Exemplarily, according to the fault type, fault severity, equipment type and production process information in the fault diagnosis result, the corresponding diagnostic instrument type and diagnostic parameter configuration are queried from the pre-established decision tree diagnostic mapping table to obtain a specific diagnostic solution. If the fault type is judged to involve rotating machinery and the fault severity reaches the preset high-risk threshold, a diagnostic instruction is sent to the on-site portable vibration diagnostic instrument through the WebSocket protocol, a fast diagnostic channel is established, and the sampling frequency of the vibration diagnostic instrument is set to 20kHz and the sampling time is 60 seconds. If the fault type is judged to involve static equipment and the fault severity does not reach the preset medium-risk threshold, the infrared thermal imager is started for diagnosis, and the infrared thermal imager parameters such as temperature measurement range, resolution and acquisition frequency are automatically adjusted according to the static equipment type. For different diagnostic instruments and fault types, an adaptive data compression algorithm is used to compress the collected diagnostic data. A wavelet transform compression algorithm is used for vibration data, and a JPEG2000 compression algorithm is used for thermal images. The compressed data is transmitted back to the central control system through a conventional data transmission channel. Based on the compressed diagnostic data, a diagnostic report containing the fault type, severity and recommended measures is automatically generated using the decision tree algorithm and sent to relevant maintenance personnel via email and text messages. According to the fault type (such as abnormal bearing wear), fault severity (such as 75%), equipment type (such as centrifugal pump) and production process (such as chemical process) information in the fault diagnosis results, the corresponding diagnostic instrument type and diagnostic parameter configuration are queried from the pre-established decision tree diagnostic mapping table to obtain a specific diagnostic plan. If the fault type is judged to involve rotating machinery, such as a centrifugal pump, and the fault severity reaches a preset high-risk threshold, such as 70%, a diagnostic instruction is sent to the on-site portable vibration diagnostic instrument through the WebSocket protocol to establish a fast diagnostic channel, and the sampling frequency of the vibration diagnostic instrument is set to 20kHz and the sampling time is set to 60 seconds to collect the radial and axial vibration data of the bearing. If the fault type is determined to involve static equipment, such as a heat exchanger, and the fault severity does not reach the preset medium-risk threshold, such as 40%, the infrared thermal imager is started for diagnosis, and the infrared thermal imager parameters are automatically adjusted according to the static equipment type, such as the temperature measurement range is set to 0-500℃, the resolution is set to 640×480 pixels, and the acquisition frequency is set to 30 frames / second. According to different diagnostic instruments and fault types, an adaptive data compression algorithm is used to compress the collected diagnostic data. For vibration data, a wavelet transform compression algorithm is used to compress the 20kHz raw data to 5kHz. For thermal images, a JPEG2000 compression algorithm is used, and the raw image compression ratio is set to 20:1. The compressed data is transmitted back to the central control system through a conventional TCP / IP data transmission channel at a rate of 256kbps.Based on the compressed diagnostic data, the C4.5 decision tree algorithm is used to automatically generate a diagnostic report containing the fault type (such as bearing inner ring failure), severity (such as 75%) and recommended measures (such as bearing replacement), and then send it to the mobile phone and email address of the relevant maintenance personnel through the SMTP protocol and GSM SMS gateway.

[0026] S104. Obtain the types of diagnostic data and data features collected by different diagnostic instruments. For vibration data collected by a portable vibration diagnostic instrument, increase the transmission priority, and obtain the corresponding transmission bandwidth allocation ratio from a preset vibration data transmission configuration table; for thermal imaging data collected by an infrared thermal imager, reduce the transmission priority, and obtain the corresponding transmission bandwidth allocation ratio from a preset thermal imaging data transmission configuration table.

[0027] The diagnostic data type and data characteristics transmitted by the diagnostic instrument data acquisition interface are obtained, and the diagnostic data types include vibration data and thermal imaging data; based on the diagnostic data type and data characteristics, the fast Fourier transform algorithm is used to calculate the spectral characteristics of the vibration data and thermal imaging data, and the time domain characteristics of the vibration data and thermal imaging data are calculated by statistics to obtain the feature vector; for the vibration data, if it is determined that its source is a portable vibration diagnostic instrument, the transmission priority of the vibration data is increased, and the corresponding transmission bandwidth allocation ratio is obtained from the preset vibration data transmission configuration table; for the thermal imaging data, if it is determined that its source is an infrared thermal imager, the transmission priority of the thermal imaging data is reduced, and the corresponding transmission bandwidth allocation ratio is obtained from the preset thermal imaging data transmission configuration table; according to the transmission priority and bandwidth allocation ratio of the vibration data and thermal imaging data, the network resources are adjusted in real time to control the transmission order and rate of the vibration data and thermal imaging data.

[0028] Exemplarily, the diagnostic data types and data features collected by different diagnostic instruments are obtained from the diagnostic instrument data acquisition interface, the frequency spectrum features of vibration data and thermal imaging data are calculated by the fast Fourier transform algorithm, and the time domain features are calculated using statistics such as mean and variance to obtain feature vectors. For the vibration data collected by the portable vibration diagnostic instrument, the priority queue algorithm is used to improve its transmission priority, and the corresponding transmission bandwidth allocation ratio is queried from the preset vibration data transmission configuration table, and the service quality mark of the network data packet is set to the highest priority using the DiffServ protocol. For the thermal imaging data collected by the infrared thermal imager, the weight allocation algorithm is used to reduce its transmission priority, and the corresponding transmission bandwidth allocation ratio is queried from the preset thermal imaging data transmission configuration table, and the service quality mark of the network data packet is set to a lower priority using the DiffServ protocol. According to the transmission priority and bandwidth allocation ratio of different data types, the weighted fair queue algorithm is used to adjust the network resources in real time, and the transmission order and rate of different types of data are controlled by the data packet scheduler to achieve optimized transmission of diagnostic data. According to the network status and data transmission situation, the adaptive bandwidth allocation algorithm is used to dynamically adjust the weight and bandwidth allocation strategy of the priority queue to adapt to different network environments and diagnostic requirements, and ensure the timely transmission of key diagnostic data. The data of the portable vibration diagnostic instrument and infrared thermal imager are obtained from the diagnostic instrument data acquisition interface, with sampling rates of 20kHz and 30Hz respectively. The spectrum of the vibration data is calculated by the 1024-point fast Fourier transform algorithm to obtain the frequency components in the range of 0-10kHz; the thermal imaging data is subjected to a two-dimensional Fourier transform to obtain the spatial frequency characteristics. The time domain characteristics such as the mean, variance, and peak factor of the vibration data, and the statistical characteristics such as the average temperature, maximum temperature, and temperature gradient of the thermal imaging data are calculated to form a feature vector. For vibration data, the weight of the priority queue algorithm is set to 0.7, and the bandwidth allocation ratio is obtained from the vibration data transmission configuration table to be 60%. The DSCP field of the IP packet is set to 46 (EF, accelerated forwarding) using the DiffServ protocol. For thermal imaging data, the weight is set to 0.3, and the bandwidth allocation ratio is obtained from the query to be 40%. The DSCP field is set to 18 (AF21, ensure forwarding). The weighted fair queue algorithm is used to allocate 60Mbps bandwidth to vibration data and 40Mbps bandwidth to thermal imaging data. The packet scheduler inserts vibration packets into the high priority queue based on the priority and bandwidth ratio, and transmits 2 thermal imaging packets for every 3 vibration packets. When the network monitoring module detects that the link utilization exceeds 85%, the adaptive bandwidth allocation algorithm increases the vibration data bandwidth to 70Mbps and reduces the thermal imaging data bandwidth to 30Mbps to ensure the timely transmission of key diagnostic data. By continuously monitoring the quality of data transmission, the priority and bandwidth allocation are dynamically adjusted to adapt to changes in the network environment.

[0029] S105. Obtain the network status and device load conditions, calculate the optimal data transmission path and server node in real time through a pre-established load balancing method, and determine the matching data transmission path and server node from the path and node mapping table based on the acquired data transmission real-time requirements and data transmission quality requirements.

[0030] The network and device state vectors are obtained, which include the network link status obtained by the network monitoring module using the simple network management protocol, and the server performance indicators obtained through the Java management extension interface; according to the network and device state vectors, they are input into a pre-established three-layer fully connected neural network model, and the output layer of the neural network model uses the Softmax function to calculate the probability distribution matrix of the optimal data transmission path and server nodes; the weight coefficients obtained by gradient descent optimization are weighted and summed with the probability distribution matrix to obtain a comprehensive score, and the transmission path and server node combination with the highest score are selected; the configuration information matching the transmission path and server node combination is searched from the path node mapping table of the hash table structure, and the configuration information includes the routing table, load balancing strategy and server resource allocation plan; the configuration information is sent to the relevant network devices and servers to complete intelligent load balancing, and the neural network model parameters and weight coefficients are continuously updated using the online learning algorithm based on the real-time network status and server load feedback.

[0031] Exemplarily, the network link status, including link bandwidth utilization, network delay, and packet loss rate, is obtained through the network monitoring module using the simple network management protocol, and the server CPU usage, memory occupancy and other indicators are obtained through the Java management extension interface to form a network and device state vector. The network and device state vectors are input into a pre-established three-layer fully connected neural network model. The input layer of the model receives the state vector, the hidden layer uses the corrected linear unit activation function, and the output layer uses the Softmax function to calculate the probability distribution matrix of the optimal data transmission path and server node in real time. According to the real-time requirements of data transmission and the quality requirements of data transmission, the weight coefficients obtained by gradient descent optimization are weighted and summed with the probability distribution matrix to obtain a comprehensive score, and the transmission path and server node combination with the highest score are selected. The matching data transmission path and server node configuration information are searched from the path node mapping table of the hash table structure, including the routing table, load balancing strategy and server resource allocation plan, and the configuration information is sent to the relevant network devices and servers to complete intelligent load balancing. According to the real-time network status and server load feedback, the neural network model parameters and weight coefficients are continuously updated using the online learning algorithm, and the load balancing strategy is dynamically adjusted to adapt to changes in the network environment and business needs. The network monitoring module uses the SNMPv3 protocol to obtain the network link status every 5 seconds, including the current value of link bandwidth utilization of 85%, the average network delay of 20ms, and the packet loss rate of 0.1%. The JMX interface is used to obtain the current value of server CPU utilization of 75% and the current value of memory occupancy of 80% every 10 seconds to form a 10-dimensional network and device state vector. The state vector is input into a pre-established three-layer fully connected neural network model. The model has 10 neurons in the input layer, 20 neurons in the hidden layer using the ReLU activation function, and 5 neurons in the output layer corresponding to 5 optional paths using the Softmax function. The probability distribution matrix of the optimal data transmission path and server nodes is calculated in real time. According to the real-time requirement of data transmission, the delay is less than 50ms, and the data transmission quality requirement, the packet loss rate is less than 0.5%, the Adam optimizer is used for gradient descent, and the learning rate is set to 0.001. The weight coefficient [0.4, 0.3, 0.3] is obtained, and the weighted sum is performed with the probability distribution matrix [0.3, 0.2, 0.2, 0.15, 0.15] to obtain the comprehensive score [0.27, 0.21, 0.21, 0.165, 0.165]. The first transmission path with the highest score and the corresponding server node are selected. The matching configuration information is searched from the path node mapping table of the hash table structure. The key is "Path1_Node3", and the value is a JSON object containing the routing table, load balancing strategy (polling algorithm) and server resource allocation plan, including a CPU limit of 75% and a memory limit of 8GB. The configuration information is sent to the relevant network devices and servers through the NETCONF protocol.Based on the real-time network status and server load feedback collected every minute, the stochastic gradient descent algorithm is used to continuously update the neural network model parameters and weight coefficients at a learning rate of 0.01. The optimal path is recalculated every hour, and the load balancing strategy is dynamically adjusted. This allows the system to quickly switch to an alternative path when the network is congested or the server is overloaded, thereby ensuring the stability of data transmission.

[0032] S106. Build a multi-level security protection system based on the data security requirements of industrial control systems and manufacturing execution systems. Encrypt the data transmitted by instant messaging with the national secret SM4, and obtain the data access rights corresponding to diagnostic personnel of different levels from the preset diagnostic personnel authority mapping table. When it is monitored that diagnostic personnel access sensitive production parameters or operate key actuators, preventive measures are taken to allow only authorized diagnostic personnel to access the corresponding level of diagnostic data and issue control instructions.

[0033] Industrial firewalls are used to divide industrial control systems and manufacturing execution systems into different security areas, and the security areas are divided according to the preset security levels. According to the division results of the security areas, the data transmitted by instant messaging is encrypted using the national secret SM4 algorithm; the preset diagnostic personnel authority mapping table is obtained, and the data access rights corresponding to diagnostic personnel of different levels are determined from the authority mapping table. According to the data access rights, diagnostic personnel are divided into three roles: operators, engineers, and administrators. The operation behavior of diagnostic personnel is monitored in real time through the anomaly detection algorithm, and the anomaly detection algorithm determines whether the operation of the diagnostic personnel is abnormal based on the preset operation behavior model. If it is detected that the diagnostic personnel access sensitive production parameters or operate key actuators, the access control list is called for permission verification. The permission information of the authorized diagnostic personnel is obtained from the access control list, and it is determined whether the diagnostic personnel have the authority to access the corresponding level of diagnostic data and issue control instructions based on the permission information.

[0034] Exemplarily, a multi-level security protection system including network isolation, identity authentication, access control, data encryption and audit log is constructed, and industrial firewalls are used to divide industrial control systems and manufacturing execution systems into different security areas, and deep packet inspection technology is used to monitor and filter inter-area communications in real time. The data transmitted by instant messaging is encrypted using the national secret SM4 algorithm, the key length is set to 128 bits, the encryption mode selects the packet link mode, and the message authentication code is used to ensure data integrity. At the same time, a key management center is established to be responsible for the secure generation, storage and distribution of SM4 keys. The data access rights corresponding to diagnostic personnel of different levels are obtained from the preset diagnostic personnel authority mapping table, and the role-based access control model is adopted to divide the diagnostic personnel into three roles: operator, engineer and administrator. The authority set of each role is defined, and the inheritance relationship between roles is established to achieve fine-grained authority control. The operation behavior of the diagnostic personnel is monitored in real time through a statistical anomaly detection algorithm. If it is detected that the diagnostic personnel access sensitive production parameters or operate key actuators, the access control list is immediately called for authority verification, and only authorized diagnostic personnel are allowed to access the corresponding level of diagnostic data and issue control instructions, and detailed operation logs are recorded at the same time. For particularly sensitive production parameters and key actuators, a two-factor authentication mechanism and operation approval process are implemented, requiring diagnostic personnel to provide dynamic passwords and biometrics for identity authentication, and only after passing multi-level approval can relevant operations be performed. In the multi-level security protection system, Huawei USG6000 series industrial firewalls are used to divide the industrial control system and manufacturing execution system into three security areas, set access control policies, and only allow authorized communication. Deep packet inspection technology uses the Snort rule library, configures 10,000 rules, monitors inter-area communications in real time, and triggers alarms when abnormal traffic is detected. Instant messaging data is encrypted using the national secret SM4 algorithm, with a key length of 128 bits, using group link mode, and updating the key once an hour. The key management center uses a hardware security module to store the master key and uses the PKCS#11 standard interface for key operations. The role-based access control model defines three roles: operator, engineer, and administrator, with 10, 20, and 30 permissions respectively, and stores the user-role mapping relationship through the LDAP directory service. The anomaly detection algorithm uses a statistical method to calculate the mean and standard deviation of user operations. When the frequency of a single operation exceeds 3 standard deviations from the mean, it is considered an anomaly. The access control list is stored in the cache, and the query response time is less than 1 millisecond. For sensitive production parameters such as reactor temperature and key actuators such as emergency shutdown valves, two-factor authentication is implemented, requiring the input of a 6-digit dynamic password and fingerprint recognition, with an identification accuracy of 99.9%. The operation approval process sets a three-level approval, and the timeout for each level of approval is 5 minutes. Centralized monitoring is carried out through the security information and event management platform, processing 1,000 logs per second to achieve real-time security situation awareness.

[0035] The above description is merely a preferred embodiment of one or more embodiments of the present specification and is not intended to limit one or more embodiments of the present specification. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principles of one or more embodiments of the present specification shall be included in the scope of protection of one or more embodiments of the present specification.

Claims

1. A method for instant messaging on a cloud communication platform, characterized in that: The method comprises: Obtain the equipment type, production process and operating parameters of industrial equipment, and combine with the pre-established equipment failure mode library to determine whether the equipment has any signs of failure. If there are signs of failure, determine the type of failure and the severity of the failure; According to the determined fault type and fault severity, the corresponding data transmission real-time requirements and data transmission quality requirements are obtained from the preset data transmission requirement mapping table. When the fault type is damage to a key component and the fault severity reaches a high-risk threshold, the video transmission resolution and frame rate are increased. When the fault type is general wear and tear and the fault severity does not reach a medium-risk threshold, the video transmission resolution and frame rate are reduced. According to the determined fault type and fault severity, as well as the equipment type and production process, the diagnostic instrument is determined from the preset diagnostic mapping table. When the fault type involves rotating machinery and the fault severity reaches the high-risk threshold, a diagnostic channel is established with the vibration diagnostic instrument through the instant messaging platform. When the fault type involves static equipment and the fault severity does not reach the medium-risk threshold, an infrared thermal imager is used for diagnosis, and the conventional data transmission channel is maintained; Obtain the diagnostic data types and data features collected by different diagnostic instruments. For vibration data collected by portable vibration diagnostic instruments, increase the transmission priority and obtain the corresponding transmission bandwidth allocation ratio from the preset vibration data transmission configuration table. For thermal imaging data collected by infrared thermal imagers, reduce the transmission priority and obtain the corresponding transmission bandwidth allocation ratio from the preset thermal imaging data transmission configuration table. Obtain network status and device load conditions, calculate the optimal data transmission path and server node in real time through a pre-established load balancing method, and determine the matching data transmission path and server node from the path and node mapping table based on the acquired data transmission real-time requirements and data transmission quality requirements; Based on the data security requirements of industrial control systems and manufacturing execution systems, a multi-level security protection system is built to encrypt data transmitted through instant messaging using the national secret SM4 encryption. The data access rights corresponding to diagnostic personnel at different levels are obtained from the preset diagnostic personnel authority mapping table. When diagnostic personnel are monitored to access sensitive production parameters or operate key actuators, blocking measures are taken to allow only authorized diagnostic personnel to access diagnostic data of the corresponding level and issue control instructions.

2. The method according to claim 1, characterized in that The device type, production process and operation parameters of the industrial equipment are obtained, and combined with a pre-established device failure mode library, it is determined whether the equipment has a failure sign, and if there is a failure sign, the failure type and the severity of the failure are determined, including: Obtain equipment type, production process and operating parameter data of industrial equipment, including real-time status information of the equipment; Preprocess the collected raw data according to the operating parameter data, including removing outliers, filling missing values ​​and data standardization, to obtain a standardized equipment operation data set; For the standardized equipment operation data set, the time series analysis method is used to calculate the statistical characteristics of various parameters, including mean, standard deviation, skewness and kurtosis, and compared with the pre-set normal parameter threshold to determine whether the equipment is in an abnormal state; If it is determined that the equipment is in an abnormal state, the abnormal state is compared with the preset fault symptom library to determine whether it is a potential fault; If it is determined to be a potential fault, the corresponding fault feature template is extracted from the pre-established equipment failure mode library according to the equipment type and production process; Use decision tree algorithm to match potential faults with fault feature templates to determine the specific fault type of the equipment; For specific fault types, a fuzzy logic controller is used to quantitatively evaluate the severity of the fault in combination with the degree of deviation and duration of equipment operating parameters. The fault is divided into three levels: minor, moderate and severe, and a fault diagnosis result report is output.

3. The method according to claim 1, characterized in that The method of obtaining the corresponding data transmission real-time requirements and data transmission quality requirements from a preset data transmission requirement mapping table according to the determined fault type and fault severity, and improving the video transmission resolution and frame rate when the fault type is damage to a key component and the fault severity reaches a high-risk threshold, and reducing the video transmission resolution and frame rate when the fault type is general wear and tear and the fault severity does not reach a medium-risk threshold, includes: Obtaining the fault type and fault severity in the fault diagnosis result from a pre-established multi-dimensional array data transmission requirement mapping table, where the multi-dimensional array data transmission requirement mapping table includes data transmission real-time requirements and data transmission quality requirements; Determine whether a critical component is damaged and reaches a preset high-risk threshold based on the fault type and fault severity; If the judgment result is that the key component is damaged and reaches the preset high-risk threshold, the video transmission resolution is increased to a preset range through the encoder, and the frame rate is increased to a preset range frame; If the judgment result is general wear and tear and the preset medium-risk threshold is not reached, the dynamic bit rate control function of the encoder is used to reduce the video transmission resolution to a preset range and the frame rate to a preset range; Dynamically adjust video encoding parameters using dynamic adaptive streaming media transmission protocol; According to the real-time network status monitoring result, a version suitable for the current network condition is selected from multiple preset bit rate versions, and the multiple bit rate versions are generated by an adaptive bit rate algorithm.

4. The method according to claim 1, characterized in that: The method includes: determining a diagnostic instrument from a preset diagnostic mapping table according to the determined fault type and fault severity as well as the equipment type and production process; establishing a diagnostic channel with a vibration diagnostic instrument through an instant messaging platform when the fault type involves rotating machinery and the fault severity reaches a high-risk threshold; and using an infrared thermal imager for diagnosis when the fault type involves static equipment and the fault severity does not reach a medium-risk threshold, and maintaining a conventional data transmission channel, including: According to the fault type, fault severity, equipment type and production process information contained in the fault diagnosis results, the corresponding diagnostic instrument type and diagnostic parameter configuration are obtained from the pre-established decision tree diagnosis mapping table to obtain a specific diagnostic plan; If the fault type is determined to involve rotating machinery and the fault severity reaches the preset high-risk threshold, a diagnostic instruction is sent to the on-site portable vibration diagnostic instrument to establish a diagnostic channel and set the sampling frequency and sampling duration of the vibration diagnostic instrument; If the fault type is determined to involve static equipment and the fault severity does not reach the preset medium-risk threshold, the infrared thermal imager is started for diagnosis and the infrared thermal imager parameters are automatically adjusted according to the static equipment type; According to the diagnostic instrument and fault type, the collected diagnostic data is compressed and transmitted back to the central control system through the data transmission channel; Based on the compressed diagnostic data, a decision tree algorithm is used to generate a diagnostic report containing the fault type, severity and recommended measures, and send it to relevant maintenance personnel.

5. The method according to claim 1, characterized in that The method of obtaining the diagnostic data types and data features collected by different diagnostic instruments, increasing the transmission priority for vibration data collected by a portable vibration diagnostic instrument, and obtaining the corresponding transmission bandwidth allocation ratio from a preset vibration data transmission configuration table, and reducing the transmission priority for thermal imaging data collected by an infrared thermal imager, and obtaining the corresponding transmission bandwidth allocation ratio from a preset thermal imaging data transmission configuration table, includes: Obtain the type of diagnostic data and data characteristics transmitted by the diagnostic instrument data acquisition interface, the diagnostic data type including vibration data and thermal imaging data; According to the type and characteristics of the diagnostic data, the fast Fourier transform algorithm is used to calculate the frequency spectrum characteristics of the vibration data and thermal imaging data, and the time domain characteristics of the vibration data and thermal imaging data are calculated through statistics to obtain the feature vector; For vibration data, if it is determined that its source is a portable vibration diagnostic instrument, the transmission priority of the vibration data is increased, and the corresponding transmission bandwidth allocation ratio is obtained from the preset vibration data transmission configuration table; For thermal imaging data, if it is determined that its source is an infrared thermal imager, the transmission priority of the thermal imaging data is reduced, and the corresponding transmission bandwidth allocation ratio is obtained from the preset thermal imaging data transmission configuration table; According to the transmission priority and bandwidth allocation ratio of vibration data and thermal imaging data, network resources are adjusted in real time to control the transmission order and rate of vibration data and thermal imaging data.

6. The method according to claim 1, characterized in that The obtaining of network status and device load conditions, calculating the optimal data transmission path and server node in real time through a pre-established load balancing method, and determining the matching data transmission path and server node from a path and node mapping table according to the acquired data transmission real-time requirements and data transmission quality requirements, includes: Obtaining network and device status vectors, including network link status obtained by the network monitoring module using the simple network management protocol, and server performance indicators obtained through the Java management extension interface; According to the network and device state vectors, they are input into the pre-established three-layer fully connected neural network model. The output layer of the neural network model uses the Softmax function to calculate the probability distribution matrix of the optimal data transmission path and server nodes. The weight coefficient is optimized by gradient descent method, and the weight coefficient is weighted and summed with the probability distribution matrix to obtain a comprehensive score, and the transmission path and server node combination with the highest score is selected; Searching for configuration information matching the transmission path and server node combination from a path node mapping table of a hash table structure, the configuration information including a routing table, a load balancing strategy, and a server resource allocation plan; The configuration information is sent to relevant network devices and servers to complete intelligent load balancing, and the neural network model parameters and weight coefficients are continuously updated using online learning algorithms based on real-time network status and server load feedback.

7. The method according to claim 1, characterized in that Based on the data security requirements of industrial control systems and manufacturing execution systems, a multi-level security protection system is constructed to encrypt the data transmitted by instant messaging with the national secret SM4, and obtain the data access rights corresponding to diagnostic personnel of different levels from the preset diagnostic personnel authority mapping table. When it is monitored that diagnostic personnel access sensitive production parameters or operate key actuators, blocking measures are taken to allow only authorized diagnostic personnel to access the corresponding level of diagnostic data and issue control instructions, including: Industrial firewalls are used to divide industrial control systems and manufacturing execution systems into different security areas, and the security areas are divided according to preset security levels; According to the division of security areas, the data transmitted by instant messaging is encrypted using the national encryption SM4 algorithm; Obtain a preset diagnostic personnel authority mapping table, and determine the data access rights corresponding to diagnostic personnel of different levels from the authority mapping table; Divide diagnostic personnel into three roles: operators, engineers, and administrators based on data access permissions; The anomaly detection algorithm is used to monitor the diagnostic personnel's operation behavior in real time. The anomaly detection algorithm determines whether the diagnostic personnel's operation is abnormal based on the preset operation behavior model; If it is detected that the diagnostic personnel access sensitive production parameters or operate key actuators, the access control list is called to perform permission verification; Obtain the permission information of the authorized diagnostic personnel from the access control list, and determine whether the diagnostic personnel have the authority to access the corresponding level of diagnostic data and issue control instructions based on the permission information.

Images