A plant-to-plant information monitoring, maintenance and management system

Abstract

The present invention relates to the field of monitoring and maintenance management systems, and provides an inter-plant information monitoring and maintenance management system; an information monitoring module is used for real-time monitoring of inter-plant information data, and sending it to a data analysis module and a decision and restriction module; the data analysis module performs production efficiency analysis, fault identification and status prediction on the data transmitted by the information detection module, and the decision and restriction module generates a user authority restriction token and an inter-plant information report, and sends it to a user operation module; the user operation module is used for displaying the inter-plant information that can be obtained by users with different authorities according to the authority restriction token and the information transmitted by the data analysis module, and sending early warning information according to the fault identification result; the present invention improves the overall planning capability of the inter-plant area through the data analysis module, effectively avoids the leakage and improper use of sensitive information through the decision and restriction module, and ensures the normal production operation of the plant area.

Description

Technical Field

[0001] The present invention relates to the field of monitoring technology and maintenance management technology, and in particular to a plant-to-plant information monitoring and maintenance management system. Background Art

[0002] With the development of industrial automation and informatization, the equipment, production process and quality control in the factory need to be monitored and managed in real time. The inter-factory information includes equipment status information, production progress information, quality data, material information, personnel information, energy consumption information and environmental monitoring information, etc. This information plays a vital role in the efficient operation, quality assurance, cost control and sustainable development of the factory.

[0003] The monitoring and maintenance management system for inter-plant information is a complex system engineering that requires the comprehensive use of multiple technologies and means. The traditional monitoring and maintenance methods for plant information usually rely on manual inspections and paper records. This method is not only inefficient, but also prone to errors, and it is difficult to adapt to the modern industrial production's demand for high efficiency, accuracy, and real-time. At present, some related technologies and products have been applied to the monitoring and management of inter-plant information, which have improved the production efficiency and management level of the factory to a certain extent.

[0004] The existing patent (patent publication number: CN108693808A) discloses an intelligent factory environment monitoring system, which comprehensively monitors the current environmental temperature, humidity, noise, air quality and parking space usage, and pushes the collected information to the display screen of each factory building, so that the staff can understand the surrounding environment information in real time; at the same time, the single-chip microcomputer comprehensively analyzes the light intensity according to the information obtained by the sensor, and switches the lighting street lights at the most reasonable time; intelligently manages parking spaces, and facilitates employees to park in an orderly and efficient manner; obtains previously delayed information in advance, so that it can quickly and accurately monitor the factory and surrounding environmental conditions, and make corresponding processing, so that managers can more conveniently understand various environmental indexes related to production operations and reasonably arrange production plans.

[0005] However, the existing background technology still has the following disadvantages:

[0006] 1. The system analysis is not comprehensive enough: Most existing systems only monitor part of the information, perform single fault identification, prediction or analysis of factory information, and cannot fully understand the various situations and information between factories, which is not conducive to improving management efficiency and overall planning between factories.

[0007] 2. Insufficient information restrictions: The existing system does not restrict the information that staff in different positions can obtain. Traditional manual management may lead to data errors or data loss, and even deliberate sabotage, which may lead to the leakage and improper use of sensitive information. Once key information is obtained or tampered with by criminals, it will mislead the judgment and decision-making of the management level, resulting in irreparable economic losses, which will have a serious impact on the production and operation of the factory area and may even cause safety accidents. Summary of the invention

[0008] The present invention provides an inter-factory information monitoring and maintenance management system, which realizes inter-factory information monitoring, analysis and maintenance management through multi-module collaboration and utilizes multiple algorithm models, and performs authority management based on personnel positions to solve the problems in the prior art.

[0009] The technical solution of the present invention is as follows:

[0010] An inter-plant information monitoring and maintenance management system, comprising:

[0011] Information monitoring module, used to monitor and receive inter-factory information data in real time, and support the sorting and transmission of data information;

[0012] A data analysis module, which is data-connected to the information monitoring module, constructs a regression model for production efficiency analysis, a multi-layer perceptron model for fault identification, and a time series algorithm for state prediction based on the data transmitted by the information monitoring module;

[0013] A decision-making and restriction module, which is data-connected with the data analysis module and the information monitoring module, and generates inter-factory information reports and permission restriction tokens according to the analysis results of the data analysis module and the information monitoring module;

[0014] The user operation module is connected to the decision and restriction module data, and based on the authority restriction token and the analysis and prediction results of the data analysis module, displays the plant-wide information analysis information that can be obtained by the corresponding authorities to users with different authorities, and sends early warning information according to the fault identification results in the data analysis module.

[0015] Preferably, the information monitoring module includes a sensor and an information collection unit;

[0016] The data information monitored and collected by the sensor includes but is not limited to environmental information, equipment status information, energy consumption information, product qualification rate data information and raw material utilization rate data information;

[0017] The information collection unit includes but is not limited to the historical production progress data information between factories and the positions of staff members based on the manually input data information.

[0018] Preferably, the data analysis module includes an analysis unit, an identification unit, a prediction unit, a maintenance management unit and an operation authority unit;

[0019] The analysis unit is used to analyze production efficiency changes and generate a production efficiency change graph;

[0020] The identification unit is used to identify the fault of the device state;

[0021] The prediction unit is used to predict the state of the device in the current normal state.

[0022] Preferably, the analysis unit constructs and trains a model through a linear regression algorithm to calculate the production efficiency, and draws a production efficiency change graph based on the production data information collected in real time by the information monitoring module:

[0023]

[0024] Wherein, x represents an independent variable, and the independent variable information includes but is not limited to energy consumption data, production progress data, product qualification rate data and raw material utilization rate data; represents the independent variable energy consumption information; Indicates the production progress information of the independent variable; Indicates the data information of the qualified rate of independent variable products; Represents the data information of the independent variable raw material utilization rate, where y represents production efficiency; represents the intercept, meaning that all independent variables When both are 0, the corresponding production efficiency is also 0. , , , Represents the regression coefficient, which indicates the influence of each independent variable on the dependent variable. The larger the regression coefficient, the deeper the influence.

[0025] Through a large amount of historical production data between factories, the mean square error is used to determine the values ​​of these regression coefficients; the error function formula is:

[0026]

[0027] Where i represents the sample index of each set of data samples, which means the i-th sample, that is, the number used to identify and distinguish each sample in the data set, and E is the sum of errors. The training process is to find the minimum value of the sum. During the training process, by adjusting the regression coefficient , , , , so that the error function is minimized and the optimal linear regression model is obtained. This process uses the least squares method to find the regression coefficients , , , To achieve:

[0028]

[0029] H is the independent variable matrix generated by different independent variables x, and is calculated based on the independent variable matrix z and the dependent variable production efficiency y , , , , T is the transpose symbol of the matrix. By transposing the independent variable matrix H and multiplying it with itself, we get a matrix , and then find its inverse matrix; then, transpose and multiply the dependent variable y and the independent variable matrix H to get a vector ; Finally, multiply the inverse matrix by this vector to get the regression coefficient β;

[0030] For each independent variable matrix, we calculate ( , , , ), constantly adjust ( , , , ) is substituted into the error function formula, and the minimum value of the error function is obtained by calculating the derivative of the error function. , , , Substitute the linear regression formula to build a linear regression model.

[0031] Preferably, the identification unit performs fault identification by a multi-layer perceptron: neurons are the basic units constituting the multi-layer perceptron;

[0032] A neuron is a function that takes multiple input features Through weighted summation and processing by activation function f, it is transformed into an output z. The multilayer perceptron is composed of multiple neurons;

[0033] For a single neuron, its function is expressed as:

[0034]

[0035] Where n is the number of samples, i is the sample index, meaning the i-th sample, are input features, is the corresponding weight, b is the bias term, and f is the ReLU activation function;

[0036] The entire multi-layer perceptron is composed of multiple neurons connected through a hierarchical structure. Different levels include input layer and output layer. The input layer receives input features, and the output layer produces the final output result.

[0037] The various parameters in the device status information monitored and collected by the sensors in the information monitoring module are used as input features , a large amount of known parameter data under normal and fault conditions is used to train the multi-layer perceptron; during the training process, the back propagation algorithm is used to continuously adjust the weights And bias term b; the back propagation algorithm consists of two steps: forward propagation and back propagation:

[0038] In the forward propagation stage, the input data is transmitted layer by layer through neurons to finally obtain the output result; in the backward propagation stage, the error between the output layer and the true value is calculated based on the actual true value obtained between the plant intervals, and the error is back-propagated back to the multi-layer perceptron to update the weights. and the bias term b, so that the trained multilayer perceptron can learn the characteristic patterns of normal state and fault state respectively; when new data is input, it is calculated and output according to the learned pattern and weight information;

[0039] The mathematical expression of the ReLU activation function is: f(z)=max(0,z);

[0040] In a multilayer perceptron, for each neuron, after calculating the output result of neuron learning, the output result is input into the ReLU function for processing; if the calculated result is greater than 0, the value is output; if the calculated result is less than or equal to 0, 0 is output;

[0041] This activation operation is performed on each neuron of the multilayer perceptron to achieve nonlinear processing and feature extraction of input data;

[0042] The output result after neuron learning is z. After being processed by the ReLU function, the output result z is 0. 0 represents a fault, indicating that the data is the same as the learned fault mode 0, and it is judged that a fault exists; otherwise, it is judged to be in a normal state.

[0043] Preferably, the prediction unit uses a time series algorithm to predict the maintenance of the equipment state in terms of the running time according to the equipment state data, and the formula is expressed as:

[0044]

[0045] in, represents each data point in the time series data, k represents the serial number of different time points r, s represents the number of historical data involved in calculating the average value, and from the time series data, select the current time r and the previous s+1 data points, add these data together, and then divide by s to get the moving average of the current time r ; Over time, this process is repeated to obtain a series of moving averages ; Ability to predict and judge future equipment status and maintenance needs based on the change patterns in the acquired historical data;

[0046] Preferably, the prediction unit judges and predicts according to the following rules:

[0047] The moving average shows a continuous upward trend, which means that the equipment is gradually deteriorating and maintenance needs to be arranged;

[0048] The moving average is stable, indicating that the equipment is currently in a stable operating state, maintenance is carried out according to the established plan, or no additional emergency maintenance is required;

[0049] The larger the change in the moving average, the more drastic the change in the equipment status, and the maintenance strategy needs to be adjusted in time; if the change is small, it is relatively stable and maintenance planning should be carried out according to the normal rhythm.

[0050] Preferably, the decision and restriction module includes a maintenance management unit and an operation authority unit;

[0051] The maintenance management unit generates a plant maintenance management plan based on the identification result of the identification unit and the prediction result of the prediction unit combined with the personnel related information;

[0052] The operation authority unit is used to generate authority restriction information and an authority restriction token according to the data information of the staff member, and the basis of the authority restriction is related to the position of the staff member.

[0053] Preferably, the operation authority unit performs authority rule matching according to personnel position information;

[0054] The operation authority unit has preset authority rules corresponding to different positions. The system matches the collected personnel position information with these preset authority rules to determine the scope and level of operation authority that the person should have. After determining the personnel's operation authority, the system generates data containing authority information; these data include but are not limited to the personnel's unique identification, employee number, authority scope, effective time and authority level;

[0055] Generate a permission restriction token based on the permission information data; verification of the permission restriction token includes but is not limited to face recognition, fingerprint verification and password.

[0056] Preferably, the user operation module includes an early warning unit, a user display interface and an authority verification unit;

[0057] The early warning unit issues an early warning according to the recognition result of the recognition unit of the data analysis module;

[0058] The user display interface is used to display the results of the analysis in the plant data analysis module, including production efficiency change graphs, fault identification results, maintenance prediction results and plant maintenance management plans;

[0059] The authority verification unit verifies the authority of the user according to the authority information data, and selectively displays operators with different authorities to the user through the user display interface.

[0060] When receiving a user's request, the permission verification unit first verifies the user's permission; if the user does not have the corresponding permission, the corresponding module data will not be displayed.

[0061] Compared with the prior art, the present invention has the following beneficial effects:

[0062] 1. The present invention comprehensively monitors various information between factories through the data analysis module. It is no longer limited to partial content and can integrate and analyze multiple types of data in the production process to realize functions such as fault identification, state prediction, production efficiency analysis and comprehensive coordination of factory information. It helps to fully understand all situations between factories and greatly improves management efficiency and the overall coordination and planning capabilities between factories.

[0063] 2. The present invention sets strict information access permission restrictions for staff in different positions through the decision-making and restriction module and the user operation module, effectively avoiding the leakage and improper use of sensitive information, and preventing key information from being illegally obtained or tampered with, thereby providing a safe and stable information environment for the production and operation of the factory, reducing the risk of safety accidents, and ensuring the normal production and operation of the factory. BRIEF DESCRIPTION OF THE DRAWINGS

[0064] Figure 1 It is a work flow chart of the present invention;

[0065] Figure 2 It is a workflow diagram of the data analysis module of the present invention. DETAILED DESCRIPTION

[0066] The following embodiments of the present invention are described in further detail in conjunction with the accompanying drawings and examples. The following examples are used to illustrate the present invention, but are not intended to limit the scope of the present invention.

[0067] The present invention provides a plant-to-plant information monitoring and maintenance management system, comprising:

[0068] Information monitoring module, used to monitor and receive inter-factory information data in real time, and support the sorting and transmission of data information;

[0069] A data analysis module, which is data-connected to the information monitoring module, constructs a regression model for production efficiency analysis, a multi-layer perceptron model for fault identification, and a time series algorithm for state prediction based on the data transmitted by the information monitoring module;

[0070] A decision-making and restriction module, which is data-connected with the data analysis module and the information monitoring module, and generates inter-factory information reports and permission restriction tokens according to the analysis results of the data analysis module and the information monitoring module;

[0071] The user operation module is connected to the decision and restriction module data, and based on the authority restriction token and the analysis and prediction results of the data analysis module, displays the plant-wide information analysis information that can be obtained by the corresponding authorities to users with different authorities, and sends early warning information according to the fault identification results in the data analysis module.

[0072] Embodiment 1: Figure 1-2 As shown, this embodiment takes a manufacturing enterprise as an example, where multiple factories are distributed in different areas. Through the inter-factory information monitoring and maintenance management system, efficient monitoring, maintenance and management of the information of each factory are achieved. The specific implementation process is as follows:

[0073] First, various sensors are placed in key locations such as production workshops, energy supply areas, and storage spaces in each factory area. The information collected by the sensors is uniformly sent to the information collection unit, and the information monitoring module processes the information uniformly;

[0074] The information monitoring module includes sensors and information collection units; advanced sensors and information collection units are installed in key locations such as production workshops, energy supply areas and storage spaces within each factory.

[0075] Sensors monitor real-time environmental information such as temperature, humidity, equipment status information, operating parameters, degree of wear, energy consumption information, energy consumption in each period, product qualification rate data information, and raw material utilization rate data information. The historical production progress data information and staff position information of the plant area are manually input into the information collection unit, and these data are transmitted to the data analysis module in real time.

[0076] At the same time, the information detection module uses the OpenRefine data cleaning and organization tool to automatically detect and convert the input data, adjust the format of the data that is detected to be non-compliant with the standard, perform unit conversion on data with inconsistent units, and trim or supplement the accuracy of data that does not meet the requirements.

[0077] The data analysis module includes an analysis unit, an identification unit, a prediction unit, a maintenance management unit, and an operation authority unit;

[0078] The analysis unit is used to analyze changes in production efficiency; a regression model is constructed using data such as energy consumption, production progress, product qualification rate, and raw material utilization rate collected from the information monitoring module as independent variables and production efficiency as the dependent variable;

[0079] The analysis unit builds and trains a model through a linear regression algorithm to calculate production efficiency. Through the production data information collected in real time by the information monitoring module, a graph of production efficiency changes is drawn:

[0080]

[0081] Wherein, x represents an independent variable, and the independent variable information includes but is not limited to energy consumption data, production progress data, product qualification rate data and raw material utilization rate data; represents the independent variable energy consumption information; Indicates the production progress information of the independent variable; Indicates the data information of the qualified rate of independent variable products; Represents the data information of the independent variable raw material utilization rate, where y represents production efficiency; represents the intercept, meaning that all independent variables When both are 0, the corresponding production efficiency is also 0. , , , Represents the regression coefficient, which indicates the influence of each independent variable on the dependent variable. The larger the regression coefficient, the deeper the influence.

[0082] Through a large amount of historical production data between factories, the mean square error is used to determine the values ​​of these regression coefficients; the error function formula is:

[0083]

[0084] Where i represents the sample index of each set of data samples, which means the i-th sample, that is, the number used to identify and distinguish each sample in the data set, and E is the sum of errors. The training process is to find the minimum value of the sum. During the training process, by adjusting the regression coefficient , , , , so that the error function is minimized and the optimal linear regression model is obtained. This process uses the least squares method to find the regression coefficients , , , To achieve:

[0085]

[0086] H is generated by arranging the independent variable values ​​corresponding to each sample in the order of columns first and rows, and is calculated based on the independent variable matrix z and the dependent variable production efficiency y. , , , , T is the transpose symbol of the matrix. By transposing the independent variable matrix H and multiplying it with itself, we get a matrix , and then find its inverse matrix; then, transpose and multiply the dependent variable y and the independent variable matrix H to get a vector ; Finally, multiply the inverse matrix by this vector to get the regression coefficient β;

[0087] For each independent variable matrix, we calculate ( , , , ), constantly adjust ( , , , ) is substituted into the error function formula, and the minimum value of the error function is obtained by calculating the derivative of the error function. , , , Substitute the linear regression formula to build a linear regression model.

[0088] Through the analysis and calculation of a large amount of data, the fluctuations of production efficiency in different time periods and the specific impact of each independent variable on production efficiency can be clearly obtained. It can also provide an important reference for fault identification.

[0089] If energy consumption is too high, it will increase energy costs, which may cause the equipment to operate in a non-optimal state, causing frequent failures and downtime for maintenance, and have a significant negative impact on production efficiency. This may mean that the equipment has an abnormal energy consumption pattern during operation, which may be a potential signal that the equipment is about to fail or has already failed. When the product qualification rate decreases and more raw materials and time are invested to achieve the production target, resulting in increased costs and a greater impact on production efficiency, it may reflect problems with the equipment's processing accuracy, thereby increasing the risk of equipment failure. When the production schedule lags and seriously affects production efficiency, it may indicate that there is a hidden danger of failure in the equipment's operating speed or stability.

[0090] By combining the analysis results of factors affecting production efficiency with equipment status parameters, a weight adjustment mechanism can be added to the fault identification model of the multi-layer perceptron. When a certain independent variable has a greater negative impact on production efficiency, the weight of the equipment status parameter related to the independent variable is correspondingly increased in fault identification, thereby more sensitively capturing possible fault signs.

[0091] The recognition unit is used to identify the fault of the equipment status, and uses a multi-layer perceptron for fault identification. The equipment status parameters collected by the sensor in the information monitoring module and the production efficiency changes calculated by the analysis unit are used as input features, and the multi-layer perceptron is trained through a large amount of known parameter data under normal and fault conditions.

[0092] The recognition unit performs fault recognition through a multi-layer perceptron: neurons are the basic units that constitute a multi-layer perceptron;

[0093] A neuron is a function that takes multiple input features Through weighted summation and processing by activation function f, it is transformed into an output z. The multilayer perceptron is composed of multiple neurons;

[0094] For a single neuron, its function is expressed as:

[0095]

[0096] Where n is the number of samples, i is the sample index, meaning the i-th sample, are input features, is the corresponding weight, b is the bias term, and f is the ReLU activation function;

[0097] The entire multi-layer perceptron is composed of multiple neurons connected through a hierarchical structure. Different levels include input layer and output layer. The input layer receives input features, and the output layer produces the final output result.

[0098] The various parameters in the device status information monitored and collected by the sensors in the information monitoring module are used as input features , a large amount of known parameter data under normal and fault conditions is used to train the multi-layer perceptron; during the training process, the back propagation algorithm is used to continuously adjust the weights And bias term b; the back propagation algorithm consists of two steps: forward propagation and back propagation:

[0099] In the forward propagation stage, the input data is transmitted layer by layer through neurons to finally obtain the output result; in the backward propagation stage, the error between the output layer and the true value is calculated based on the true value actually obtained between the plant intervals, that is, the accurate labeling of known normal or faulty values, and the error is back-propagated back to the multi-layer perceptron to update the weights. and bias term b, so that the model is gradually optimized, and the characteristic laws and patterns in the normal state and fault state data are captured according to the optimized different weights and bias terms representing the normal state or fault state. When the equipment state parameters collected by the sensors in the new information monitoring module and the production efficiency change values ​​calculated by the analysis unit are input, the calculation and output are performed according to the learned pattern and weight information;

[0100] Among them, the mathematical expression of the ReLU activation function is: f(z)=max(0,z);

[0101] In a multilayer perceptron, for each neuron, after calculating the output result of neuron learning, the output result is input into the ReLU function for processing; if the calculated result is greater than 0, the value is output; if the calculated result is less than or equal to 0, 0 is output;

[0102] This activation operation is performed on each neuron of the multilayer perceptron to achieve nonlinear processing and feature extraction of input data;

[0103] The output result after neuron learning is z. The ReLU function is a recognition function that finds the larger value of z and 0. When the output result z is 0 after being processed by the ReLU function, because the calculated z is always greater than 0 when the equipment between the factory areas is working normally, 0 represents a fault, and z is 0, which means that a fault is judged to exist.

[0104] During operation, whenever new equipment status data is input, the system can quickly and accurately determine whether the equipment is faulty based on the learned multi-layer perceptron. When the operating parameters of a key production equipment fluctuate abnormally, the system promptly identifies the risk of failure of the equipment and issues an alarm.

[0105] The prediction unit is used to predict the status of the equipment in the current normal state; based on but not limited to equipment status data, fault identification results and changes in production efficiency, use time series algorithms to predict the maintenance of equipment status, select the current moment and multiple previous data points from historical data, and calculate the moving average; according to the changing trend of the moving average, such as continuous rise, stability, and the magnitude of the change, accurately judge the development trend of the equipment status.

[0106] The prediction unit uses the time series algorithm to predict the maintenance of the equipment status in terms of the running time according to the equipment status data. The formula is expressed as:

[0107]

[0108] in, represents each data point in the time series data, k represents the serial number of different time points r, s represents the number of historical data involved in calculating the average value, and from the time series data, select the current time r and the previous s+1 data points, add these data together, and then divide by s to get the moving average of the current time r ; Over time, this process is repeated to obtain a series of moving averages ; Ability to predict and judge future equipment status and maintenance needs based on the change patterns in the acquired historical data;

[0109] The prediction unit judges and predicts according to the following rules:

[0110] The moving average shows a continuous upward trend, which means that the equipment is gradually deteriorating and maintenance needs to be arranged;

[0111] The moving average is stable, indicating that the equipment is currently in a stable operating state, maintenance is carried out according to the established plan, or no additional emergency maintenance is required;

[0112] The larger the change in the moving average, the more drastic the change in the equipment status, and the maintenance strategy needs to be adjusted in time; if the change is small, it is relatively stable and maintenance planning should be carried out according to the normal rhythm.

[0113] When the moving average of a certain device shows a continuous upward trend, the system predicts that the device is gradually deteriorating, indicating that urgent maintenance is needed; for equipment with a stable moving average, it indicates that it is currently in a stable operating state and does not require urgent maintenance, and maintenance can be carried out according to the original factory maintenance schedule.

[0114] Existing technologies often only focus on a single data analysis or fault detection function; this solution collects data comprehensively through the information monitoring module, and the various algorithm models of the data analysis module deeply mine the data value. The results of the analysis unit provide a basis for the identification unit to adjust the weight of the equipment status parameters, which helps to capture fault signs more keenly; the results of the analysis unit and the identification unit provide support for the prediction unit to judge the development trend of the equipment status, thereby jointly providing strong support for the comprehensive management of the plant and realizing comprehensive monitoring, optimizing maintenance strategies and ensuring stable production. It not only realizes the integration of production efficiency analysis, fault identification and status prediction functions, but also realizes the deep integration and collaborative application of data from different sources and types, thus providing strong support for the comprehensive management of the plant and providing a comprehensive and accurate decision-making basis for the comprehensive management of the plant and providing a comprehensive understanding of all situations between the plant and greatly improving management efficiency and the overall planning capabilities of the plant and between the plants.

[0115] Embodiment 2: Figure 1-2As shown, in the plant area of ​​the first embodiment, relevant personnel need to check the relevant plant area maintenance management plan through the system;

[0116] The decision and restriction module includes a maintenance management unit and an operation authority unit;

[0117] The maintenance management unit generates a detailed and scientific maintenance management plan for each plant area based on the results of the identification unit and the prediction unit, combined with relevant information such as the staff's shift schedule and skill level. For equipment that is about to require maintenance, maintenance personnel with corresponding skills and experience are reasonably arranged, and the specific time for maintenance and the required tools and spare parts are determined. These data are manually input by the information collection unit;

[0118] The operation permission unit is used to generate permission restriction information and permission restriction tokens according to the data information of the staff. The basis of permission restriction is related to the position of the staff.

[0119] The operation authority unit has preset authority rules corresponding to different positions. The system matches the collected personnel position information with these preset authority rules to determine the scope and level of operation authority that the person should have. After determining the personnel's operation authority, the system generates data containing authority information; these data include but are not limited to the personnel's unique identification, employee number, authority scope, effective time and authority level;

[0120] Generate a permission restriction token based on the permission information data; verification of the permission restriction token includes but is not limited to face recognition, fingerprint verification and password.

[0121] The user operation module includes an early warning unit, a user display interface and an authority verification unit;

[0122] The early warning unit issues an early warning according to the recognition result of the recognition unit of the data analysis module;

[0123] The user display interface is used to display the analysis results in the plant data analysis module, including production efficiency change graphs, fault identification results, maintenance prediction results, and plant maintenance management plans;

[0124] The authority verification unit verifies the authority of the user according to the authority information data, and selectively displays operators with different authorities to the user through the user display interface.

[0125] When receiving a user's request, the permission verification unit first verifies the user's permission; if the user does not have the corresponding permission, the corresponding module data will not be displayed.

[0126] Frontline workers can see detailed maintenance management plans, operation guides and precautions related to the equipment they are responsible for on the user display interface; workshop supervisors can view the equipment status, production efficiency data and maintenance plan execution progress of the entire workshop; senior managers can fully understand the comprehensive data and overall operation status of multiple plant areas on the user display interface. The authority verification unit strictly verifies the user's access rights to ensure the security and confidentiality of data. Only verified users can obtain and operate the corresponding data and functions according to their own permissions.

[0127] Most of the existing rights management systems are relatively simple and cannot perform fine division of rights and dynamic adjustment according to the specific positions and work requirements of the factory staff. The operation permission unit in this solution can flexibly assign corresponding operation permissions to different users according to the personnel's position information and work scenarios, ensuring the security and confidentiality of data, while improving the efficiency of system use and management standardization, providing a safe and stable information environment for the production and operation of the factory, reducing the risk of safety accidents, and ensuring the normal production and operation of the factory.

[0128] The embodiments of the present invention are provided for the purpose of illustration and description. Although the embodiments of the present invention have been shown and described above, it can be understood that the above embodiments are exemplary and cannot be understood as limitations of the present invention. Ordinary technicians in this field can change, modify, replace and modify the above embodiments within the scope of the present invention.

Claims

1. An inter-factory information monitoring and maintenance management system, characterized in that: include: Information monitoring module, used to monitor and receive inter-factory information data in real time, and support the sorting and transmission of data information; A data analysis module, which is data-connected to the information monitoring module, constructs a regression model for production efficiency analysis, a multi-layer perceptron model for fault identification, and a time series algorithm for state prediction based on the data transmitted by the information monitoring module; A decision-making and restriction module, which is data-connected with the data analysis module and the information monitoring module, and generates inter-factory information reports and permission restriction tokens according to the analysis results of the data analysis module and the information monitoring module; A user operation module, which is data-connected with the decision and restriction module, and displays the plant-wide information analysis information that can be obtained by users with different permissions based on the permission restriction token and the analysis and prediction results of the data analysis module, and sends warning information according to the fault identification results in the data analysis module; The data analysis module includes an analysis unit, an identification unit, a prediction unit, a maintenance management unit and an operation authority unit; The analysis unit is used to analyze production efficiency changes and generate a production efficiency change graph; The identification unit is used to identify the fault of the device state; The prediction unit is used to predict the state of the device in the current normal state; The analysis unit constructs and trains a model through a linear regression algorithm to calculate production efficiency, and draws a production efficiency change graph through production data information collected in real time by the information monitoring module: Wherein, x represents an independent variable, and the independent variable information includes but is not limited to energy consumption data, production progress data, product qualification rate data and raw material utilization rate data; represents the independent variable energy consumption information; Indicates the production progress information of the independent variable; Indicates the data information of the qualified rate of independent variable products; Represents the data information of the independent variable raw material utilization rate, where y represents production efficiency; represents the intercept, meaning that all independent variables When both are 0, the corresponding production efficiency y is also 0. , , , Represents the regression coefficient, which indicates the influence of each independent variable on the dependent variable. The larger the regression coefficient, the deeper the influence. Through a large amount of historical production data between factories, the mean square error is used to determine the values ​​of these regression coefficients; The error function formula is: Where i represents the sample index of each group of data samples, That is the independent variable data samples The calculated production efficiency is the number used to identify and distinguish each sample in the data set. E is the sum of errors. The training process is to find the minimum value of the sum. During the training process, by adjusting the regression coefficient , , , , so that the error function is minimized, the optimal linear regression model is obtained, and the least squares method is used to find the regression coefficient in this process of minimizing the error function. , , , To achieve this, the formula is: H is the independent variable matrix generated by different independent variables x, and is calculated based on the independent variable matrix z and the dependent variable production efficiency y , , , , T is the transpose symbol of the matrix. By transposing the independent variable matrix H and multiplying it with itself, we get a matrix , and then find its inverse matrix; then, transpose and multiply the dependent variable production efficiency y and the independent variable matrix H to obtain a vector ; Finally, multiply the inverse matrix by this vector to get the regression coefficient β; For each independent variable matrix, we calculate ( , , , ), constantly adjust ( , , , ) is substituted into the error function formula, and the minimum value of the error function is obtained by calculating the derivative of the error function. , , , Substitute the linear regression formula to build a linear regression model.

2. The inter-plant information monitoring and maintenance management system according to claim 1, characterized in that: The information monitoring module includes a sensor and an information collection unit; The data information monitored and collected by the sensor includes but is not limited to environmental information, equipment status information, energy consumption information, product qualification rate data information and raw material utilization rate data information; The information collection unit includes but is not limited to the historical production progress data information between factories and the positions of staff members based on the manually input data information.

3. The inter-plant information monitoring and maintenance management system as claimed in claim 1, characterized in that: The identification unit performs fault identification through a multi-layer perceptron: neurons are the basic units constituting the multi-layer perceptron; A neuron is a function that takes multiple input features Through weighted summation and processing by activation function f, it is transformed into an output z. The multilayer perceptron is composed of multiple neurons; For a single neuron, its function is expressed as: Where n is the number of samples, i is the sample index, meaning the i-th sample, are input features, is the corresponding weight, b is the bias term, and f is the ReLU activation function; The entire multi-layer perceptron is composed of multiple neurons connected through a hierarchical structure. Different levels include input layer and output layer. The input layer receives input features, and the output layer produces the final output result. The various parameters in the device status information monitored and collected by the sensors in the information monitoring module are used as input features , a large amount of known parameter data under normal and fault conditions is used to train the multi-layer perceptron; during the training process, the back propagation algorithm is used to continuously adjust the weights And bias term b; the back propagation algorithm consists of two steps: forward propagation and back propagation: In the forward propagation stage, the input data is transmitted layer by layer through neurons to finally obtain the output result; in the backward propagation stage, the error between the output layer and the true value is calculated according to the true value of the fault result actually obtained between the plant and the inter-plant area, and the error is back-propagated back to the multi-layer perceptron to update the weights. and the bias term b, so that the trained multilayer perceptron can learn the characteristic patterns of normal state and fault state respectively; when new data is input, it is calculated and output according to the learned pattern and weight information; The mathematical expression of the ReLU activation function is: f(z)=max(0,z); In a multilayer perceptron, for each neuron, after calculating the output result of neuron learning, the output result is input into the ReLU function for processing; if the calculated result is greater than 0, the value is output; if the calculated result is less than or equal to 0, 0 is output; The above activation operation is performed on each neuron of the multilayer perceptron to achieve nonlinear processing and feature extraction of input data; The output result after neuron learning is z. After being processed by the ReLU function, the output result z is 0. 0 represents a fault, indicating that the data is the same as the learned fault mode 0, and it is judged that a fault exists; otherwise, it is judged to be in a normal state.

4. The inter-plant information monitoring and maintenance management system as claimed in claim 1, characterized in that: The prediction unit uses a time series algorithm to predict the maintenance of the equipment status in terms of the running time according to the equipment status data. The formula is expressed as: in, represents each data point in the time series data, k represents the serial number of different time points r, s represents the number of historical data involved in calculating the average value, and from the time series data, select the current time r and the previous s+1 data points, add these data together, and then divide by s to get the moving average of the current time r ; Over time, this process is repeated to obtain a series of moving averages ; It can predict and judge the future equipment status and maintenance needs based on the change patterns in the historical data obtained.

5. The inter-plant information monitoring and maintenance management system as claimed in claim 4, characterized in that: The prediction unit judges and predicts according to the following rules: The moving average shows a continuous upward trend, which means that the equipment is gradually deteriorating and maintenance needs to be arranged; The moving average is stable, indicating that the equipment is currently in a stable operating state, maintenance is carried out according to the established plan, or no additional emergency maintenance is required; The larger the change in the moving average, the more drastic the change in the equipment status, and the maintenance strategy needs to be adjusted in time; if the change is small, it is relatively stable and maintenance planning should be carried out according to the normal rhythm.

6. The inter-plant information monitoring and maintenance management system as claimed in claim 5, characterized in that: The decision and restriction module includes a maintenance management unit and an operation authority unit; The maintenance management unit generates a plant maintenance management plan based on the identification result of the identification unit and the prediction result of the prediction unit combined with the personnel related information; The operation authority unit is used to generate authority restriction information and an authority restriction token according to the data information of the staff member, and the basis of the authority restriction is related to the position of the staff member.

7. The inter-plant information monitoring and maintenance management system as claimed in claim 6, characterized in that: The operation authority unit performs authority rule matching according to personnel position information; The operation authority unit has preset authority rules corresponding to different positions. The system matches the collected personnel position information with these preset authority rules to determine the scope and level of operation authority that the person should have. After determining the personnel's operation authority, the system generates data containing authority information; these data include but are not limited to the personnel's unique identification, employee number, authority scope, effective time and authority level; Generate a permission restriction token based on permission information data; Verification of permission restriction tokens includes but is not limited to face recognition, fingerprint verification and password.

8. The inter-plant information monitoring and maintenance management system as claimed in claim 7, characterized in that: The user operation module includes an early warning unit, a user display interface and an authority verification unit; The early warning unit issues an early warning according to the recognition result of the recognition unit of the data analysis module; The user display interface is used to display the results of the analysis in the plant data analysis module, including production efficiency change graphs, fault identification results, maintenance prediction results and plant maintenance management plans; The authority verification unit verifies the authority of the user according to the authority information data, and selectively displays operators with different authorities to the user through the user display interface; When receiving a user's request, the permission verification unit first verifies the user's permission; if the user does not have the corresponding permission, the corresponding module data will not be displayed.

Images