Chemical production control system

By integrating data acquisition, processing and diagnosis modules in the chemical production control system, using high-precision sensors and recurrent neural networks for fault diagnosis and early warning, the problem of insufficient fault diagnosis capabilities of traditional systems is solved, and the safety and efficiency of chemical production is achieved promote.

CN119292221AInactive Publication Date: 2025-01-10NANTONG CHAOYUAN NEW MATERIAL CO LTD
View PDF 0 Cites 3 Cited by

Patent Information

Application Number
CN202411641844.4
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2024-11-18
Publication Date
2025-01-10
Estimated Expiration
Not applicable · inactive patent

AI Technical Summary

Technical Problem

Traditional chemical production control systems have limited capabilities in fault diagnosis and early warning, resulting in production interruptions and safety hazards, which require manual inspection and repair, which is time-consuming and labor-intensive.

Method used

A chemical production control system including data acquisition, data processing, fault diagnosis and early warning modules was designed. Data is collected in real time through multiple fault diagnosis models and high-precision sensors, and data processing is performed using filtering, denoising and feature extraction algorithms. Use recurrent neural network to automatically diagnose faults and issue early warning signals in a timely manner.

Benefits of technology

It realizes accurate identification and rapid positioning of different types of faults, prevents the expansion of faults, improves the safety and production efficiency of chemical production, and reduces production costs.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN119292221A_ABST
    Figure CN119292221A_ABST
Patent Text Reader

Abstract

The invention discloses a chemical production control system, and relates to the technical field of chemical production automation, the system comprises the following components: a data acquisition module, a data processing module, a fault diagnosis module and an early warning module, the data acquisition module is used for acquiring various data in a chemical production process in real time; the data processing module is used for processing and analyzing the collected data; the fault diagnosis module is used for performing automatic fault diagnosis based on the processed data; according to the invention, through a plurality of fault diagnosis models in the fault diagnosis module, different types of faults can be accurately identified, and the system can rapidly position and give out early warning no matter equipment faults, process abnormalities or operation errors, and in addition, the fault diagnosis accuracy is improved. The early warning module can send early warning signals in multiple modes, it is ensured that an operator can rapidly receive information and take corresponding countermeasures, and the rapid response mechanism is beneficial for preventing fault expansion and further guaranteeing safe proceeding of chemical production.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The invention relates to the technical field of chemical production automation, in particular to a chemical production control system. Background Art

[0002] At present, with the gradual expansion of the production scale and achievable functions of chemical enterprises, the scale and complexity of the production equipment configured for chemical production have also been greatly improved. In this environment, relying solely on manual production control can no longer meet the production requirements of chemical enterprises. The development of information technology and its application in many fields have made chemical production automation more and more popular, driving the product production control of chemical enterprises to gradually develop in the direction of automation, intelligence, scale and complexity.

[0003] Traditional chemical production control systems have many shortcomings in monitoring and fault handling. Although they can monitor some basic production parameters, their ability to diagnose and warn of complex faults is limited. Once a fault occurs, it is usually necessary to rely on manual troubleshooting and repair, which is not only time-consuming and labor-intensive, but may also lead to production interruptions and even safety accidents. Therefore, developing a chemical production control system that can automatically diagnose faults and provide early warnings is of great significance to improving the overall level and safety of chemical production. Summary of the invention

[0004] The purpose of the present invention is to make up for the shortcomings of the prior art and to provide a chemical production control system, which can accurately identify different types of faults through a variety of fault diagnosis models in the fault diagnosis module. Whether it is equipment failure, process abnormality or operational error, the system can quickly locate and issue an early warning. In addition, the early warning module can send early warning signals in a variety of ways to ensure that operators can quickly receive information and take corresponding countermeasures. This rapid response mechanism helps prevent the expansion of faults and further ensure the safe progress of chemical production.

[0005] In order to solve the above technical problems, the present invention provides the following technical solutions: a chemical production control system, comprising a data acquisition module, a data processing module, a fault diagnosis module and an early warning module, wherein the data acquisition module is used to collect various data in the chemical production process in real time; the data processing module is used to process and analyze the collected data; the fault diagnosis module is used to automatically diagnose faults based on the processed data; and the early warning module is used to promptly issue an early warning signal after diagnosing a fault.

[0006] Furthermore, the data acquisition module includes a variety of high-precision sensors and signal conditioning circuits for real-time acquisition of a variety of parameters in the chemical production process, and amplification and filtering of the acquired signals to improve signal quality and stability.

[0007] Furthermore, the data processing module uses advanced data processing algorithms, including filtering, denoising, standardization preprocessing operations, and feature extraction algorithms to extract key features related to the fault, providing strong support for subsequent fault diagnosis:

[0008] (1) Filtering algorithm:

[0009] The filtering algorithm is used to eliminate noise and interference in the signal. The formula is: Where (x[n]) is the input signal, (y[n]) is the filtered output signal, (n) is the window size of the moving average, and the implementation steps are: determine the window size (n), traverse the input signal (x[n]), and for each sampling point calculate the average value of the previous (n) sampling points as the output;

[0010] (2) Mean filtering denoising:

[0011] The mean filter replaces the value of each data point with the average value of its surrounding neighboring data points. The formula is: Where (y[n]) is the filtered output, (x[n]) is the original signal, (S) is the selected neighborhood, (m) is the total number of pixels or data points in the neighborhood, and the denoising algorithm is used to further reduce the noise in the signal;

[0012] Feature extraction algorithm:

[0013] The feature extraction algorithm is used to extract key features related to the fault from the preprocessed signal. The algorithms involved include peak detection, mean calculation, and variance calculation.

[0014] Peak detection: Identify peaks in a signal by comparing the values ​​of adjacent sampling points;

[0015] Mean calculation: Calculate the average value of the signal to reflect the overall level of the signal. The calculation formula is: Where (n) is the number of data points, (x i ) is the value of each data point, and the formula is used to calculate the average of all values ​​in the data set;

[0016] Variance calculation: Calculate the average of the square of the deviation between the signal value and the mean, which reflects the degree of fluctuation of the signal. The calculation formula of the variance is: Where (n) is the number of data points, (x i ) is the value of each data point, and the mean is the average of the data set. This formula averages the squares of the differences between each data point and the mean, providing a measure of the dispersion of the data distribution.

[0017] Furthermore, the fault diagnosis module implements the following steps:

[0018] (1) Data collection and preprocessing: Collect relevant chemical production process data and perform cleaning and normalization preprocessing operations;

[0019] (2) Feature extraction: extracting meaningful features from the preprocessed data, which will be used for subsequent fault diagnosis;

[0020] (3) Model training: Use the selected algorithm and collected data to train the fault diagnosis model;

[0021] (4) Model evaluation and optimization: Evaluate the performance of the model and adjust parameters or optimize the model as needed;

[0022] (5) Online fault diagnosis: Deploy the trained model to the actual system for real-time detection and diagnosis of faults.

[0023] Furthermore, the normalized calculation formula is as follows:

[0024] Z-score normalization, also known as standardization, is a method of mapping data to a normal distribution with a mean of 0 and a standard deviation of 1. The specific calculation formula is as follows:

[0025]

[0026] Among them, x norm is the normalized data, (x) is the original data, (μ) is the mean of the original data, and (σ) is the standard deviation of the original data. This normalization method can eliminate the scale difference of the data so that different features or variables have the same scale.

[0027] Furthermore, the model training adopts a recurrent neural network (RNN) algorithm, and the formula of RNN can be expressed as: ht=f(W·ht-1+U·xt+b)h t =f(W·h t-1 +U·x t +b)ht=f(W·ht-1+U·xt+b), where: ht represents the hidden state at the tth moment, W and U are weight matrices used to capture the correlation between the input and hidden state at different moments, b is the bias term, f is the activation function used to introduce nonlinear factors so that the model can capture complex patterns, and xt is the input at the tth moment.

[0028] This formula reflects the recurrent connection in RNN, that is, the output at each moment depends not only on the input at the current moment, but also on the hidden state at the previous moment. This structure enables RNN to capture the temporal dependency and correlation in sequence data, making it suitable for processing time series, speech or text sequence data.

[0029] Furthermore, the data acquisition module collects various data in the chemical production process in real time; the data processing module processes and analyzes the collected data to extract key features related to the fault; the fault diagnosis module automatically diagnoses the fault based on the processed data; after the fault is diagnosed, the early warning module promptly issues an early warning signal so that the operator can take timely measures to deal with the fault.

[0030] Furthermore, the data acquisition module can collect temperature, pressure, flow, and liquid level parameters in real time through a variety of high-precision sensors so as to collect various data in the chemical production process.

[0031] Furthermore, the fault diagnosis module has built-in multiple fault diagnosis models, including rule-based diagnosis models and machine learning-based diagnosis models, which can automatically diagnose faults in the chemical production process based on processed data and extracted features.

[0032] Furthermore, the early warning module has multiple early warning modes, including sound and light alarms, SMS notifications, and email reminders, which can ensure the timeliness and effectiveness of early warning information.

[0033] 1. The present invention can accurately identify different types of faults through various fault diagnosis models in the fault diagnosis module. Whether it is equipment failure, process abnormality or operational error, the system can quickly locate and issue an early warning. In addition, the early warning module can send early warning signals in a variety of ways to ensure that operators can quickly receive information and take corresponding countermeasures. This rapid response mechanism helps prevent the expansion of faults and further ensure the safe progress of chemical production.

[0034] 2. Through real-time monitoring and data analysis, the present invention enables operators to better understand the bottlenecks and potential problems in the production process, thereby optimizing the production process and improving production efficiency. At the same time, since faults can be discovered and handled in a timely manner, production interruptions and equipment damage caused by faults are avoided, further reducing production costs.

[0035] Other advantages, objectives and features of the present invention will be set forth in part in the following description and, in part, will be apparent to those skilled in the art based on an examination of the following or may be taught from the practice of the present invention. BRIEF DESCRIPTION OF THE DRAWINGS

[0036] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments or the prior art descriptions are briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention, and for ordinary technicians in this field, other drawings can be obtained based on these drawings without creative work.

[0037] Figure 1 This is a flow chart of a chemical production control system. DETAILED DESCRIPTION

[0038] The technical solutions in the embodiments of the present invention are described clearly and completely below. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of them. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0039] Example 1: Data collection and preprocessing

[0040] In this embodiment, we describe in detail the data collection and preprocessing process in a chemical production control system.

[0041] First, the data acquisition module uses multiple sensors installed on chemical production equipment to collect key parameters such as temperature, pressure, flow, and liquid level in real time. These sensors transmit data to the data acquisition module via wired or wireless means.

[0042] Next, the data acquisition module will preliminarily organize the received raw data, remove outliers and duplicate data, and sort the data in chronological order.

[0043] Then, the data processing module uses advanced data processing technology to filter and denoise the raw data to eliminate sensor noise and interference signals. The mean filter replaces the value of each data point with the average value of its surrounding neighboring data points. The formula is: Where (y[n]) is the filtered output, (x[n]) is the original signal, (S) is the selected neighborhood, (m) is the total number of pixels or data points in the neighborhood, and a denoising algorithm is used to further reduce the noise in the signal.

[0044] Finally, the data processing module extracts key features related to the fault from the filtered data through feature extraction technology. These features can be certain statistics, waveform parameters, or features extracted based on machine learning algorithms.

[0045] Through the above steps, the data acquisition and preprocessing module completes the collection and preliminary processing of chemical production data, providing accurate and reliable data support for subsequent fault diagnosis.

[0046] Embodiment 2: Automatic fault diagnosis

[0047] In this embodiment, the implementation process of automatic fault diagnosis in a chemical production control system is introduced in detail.

[0048] First, the fault diagnosis module receives the processed data from the data processing module. These data have been filtered, denoised and feature extracted, and have high quality and relevance.

[0049] Next, the fault diagnosis module uses a variety of built-in fault diagnosis models to perform automatic fault diagnosis. These models can be rule-based models, statistical models, or machine learning-based models. The specific model selected depends on the characteristics of the chemical production process and the type of fault.

[0050] For models based on machine learning, the fault diagnosis module will use historical fault data for model training. Through continuous iteration and optimization, the model can learn the inherent laws and characteristics of the fault data, thereby achieving accurate diagnosis of new data.

[0051] During the diagnosis process, the fault diagnosis module will classify or regress the input data to determine whether a fault exists and the type and extent of the fault. If a fault is diagnosed, the module will record the fault information and prepare to trigger the early warning module.

[0052] Through the application of automatic fault diagnosis modules, chemical production control systems can achieve rapid and accurate fault diagnosis, thus improving the safety and efficiency of the production process.

[0053] Example 3: Triggering and sending warning signals

[0054] In this embodiment, we describe the implementation process of triggering and sending early warning signals in a chemical production control system.

[0055] Once the fault diagnosis module diagnoses a fault, the early warning module will be activated immediately. The early warning module will first record the fault information, including the fault type, occurrence time and severity.

[0056] Next, the warning module will determine whether to trigger a warning signal based on the preset warning rules. The warning rules can be flexibly set according to actual conditions. For example, a warning may be triggered when the severity of the fault exceeds a certain threshold.

[0057] Once the warning signal is triggered, the warning module will select an appropriate sending method to send the warning information to the operator in a timely manner. The sending methods may include sound and light alarms, SMS notifications, and email reminders. Sound and light alarms can be achieved through the on-site alarm device to promptly remind nearby operators; SMS notifications and email reminders can be sent to remote operators through mobile communication networks or the Internet to ensure that they can obtain fault information in a timely manner.

[0058] The content of the early warning signal should contain the key information of the fault so that the operator can quickly understand the fault situation and take corresponding treatment measures. At the same time, the early warning module can also provide a query function for the fault history record to facilitate the operator to analyze and summarize the fault.

[0059] Through the application of early warning signal triggering and sending modules, the chemical production control system can achieve rapid response and timely processing of faults, effectively improving the safety and efficiency of the production process.

[0060] Example 4: Application of fault diagnosis algorithm based on deep learning

[0061] In this example, we demonstrated how to apply a deep learning-based fault diagnosis algorithm to a chemical production control system.

[0062] First, we collect a large amount of historical fault data as training samples. These data include various parameters in the chemical production process and the corresponding fault labels.

[0063] Next, we build a deep learning model by using a recurrent neural network (RNN) with the formula: ht = f(W·ht-1+U·xt+b)h t =f(W·h t-1 +U·x t +b)ht=f(W·ht-1+U·xt+b), where: ht represents the hidden state at the tth moment, W and U are weight matrices used to capture the correlation between the input and hidden state at different moments, b is the bias term, f is the activation function used to introduce nonlinear factors so that the model can capture complex patterns, and xt is the input at the tth moment. The above algorithm can be used to learn the inherent laws and characteristics of fault data, and the input of the model can be raw data or a feature vector that has been preliminarily processed.

[0064] Then, we use the training samples to train the deep learning model. Through the back-propagation algorithm and gradient descent optimization, the model can gradually learn the feature representation and classification boundaries of the fault data. The basic steps of the back-propagation algorithm include:

[0065] Forward propagation: calculate the output value of each layer;

[0066] Calculate the error of the output layer: Calculate the error of the output layer according to the loss function;

[0067] Back propagation error: Use the chain rule to calculate the error of each layer layer by layer;

[0068] Calculate gradients: Calculate the gradient of the loss function with respect to the parameters based on the error of each layer and the derivative of the activation function;

[0069] In addition, gradient descent optimization is an optimization algorithm for updating model parameters. Its basic idea is to update the parameters in the opposite direction of the gradient of the loss function so that the loss function gradually decreases. Specifically, the formula of the gradient descent algorithm is:

[0070]

[0071] Among them, θ is the model parameter, α is the learning rate (controls the step size of each update), J(θ) is the loss function, It is the gradient of the loss function with respect to the parameters. Through continuous iteration of this process, the model parameters will gradually converge to the optimal solution that minimizes the loss function.

[0072] After the model training is completed, we can deploy it to the fault diagnosis module of the chemical production control system. When new production data is input, the model will classify or regress it to determine whether there is a fault and the type and degree of the fault.

[0073] The fault diagnosis algorithm based on deep learning has powerful feature learning and generalization capabilities, and can handle complex nonlinear relationships and uncertain factors. By applying this algorithm, the chemical production control system can accurately diagnose faults and improve the accuracy and reliability of fault detection.

[0074] It will be apparent to those skilled in the art that the invention is not limited to the details of the exemplary embodiments described above and that the invention can be implemented in other specific forms without departing from the spirit or essential features of the invention. Therefore, the embodiments should be considered in all respects as exemplary and non-limiting, and the scope of the invention is defined by the appended claims rather than the foregoing description, and it is intended that all variations falling within the meaning and scope of the same elements of the claims be included in the invention. Any reference numeral in a claim should not be considered as limiting the claim to which it relates.

Claims

1. A chemical production control system, characterized in that: It includes a data acquisition module, a data processing module, a fault diagnosis module and an early warning module. The data acquisition module is used to collect various data in the chemical production process in real time; the data processing module is used to process and analyze the collected data; the fault diagnosis module is used to automatically diagnose faults based on the processed data; and the early warning module is used to issue an early warning signal in time after diagnosing a fault.

2. A chemical production control system according to claim 1, characterized in that: The data acquisition module includes a variety of high-precision sensors and signal conditioning circuits, which are used to collect a variety of parameters in the chemical production process in real time, and amplify and filter the collected signals to improve the quality and stability of the signals.

3. A chemical production control system according to claim 1, characterized in that: The data processing module uses advanced data processing algorithms, including filtering, denoising, standardization preprocessing operations, and feature extraction algorithms to extract key features related to the fault, providing strong support for subsequent fault diagnosis: (1) Filtering algorithm: The filtering algorithm is used to eliminate noise and interference in the signal. The formula is: Where (x[n]) is the input signal, (y[n]) is the filtered output signal, (n) is the window size of the moving average, and the implementation steps are: determine the window size (n), traverse the input signal (x[n]), and for each sampling point calculate the average value of the previous (n) sampling points as the output; (2) Mean filtering denoising: The mean filter replaces the value of each data point with the average value of its surrounding neighboring data points. The formula is: Where (y[n]) is the filtered output, (x[n]) is the original signal, (S) is the selected neighborhood, (m) is the total number of pixels or data points in the neighborhood, and the denoising algorithm is used to further reduce the noise in the signal; Feature extraction algorithm: The feature extraction algorithm is used to extract key features related to the fault from the preprocessed signal. The algorithms involved include peak detection, mean calculation, and variance calculation. Peak detection: Identify peaks in a signal by comparing the values ​​of adjacent sampling points; Mean calculation: Calculate the average value of the signal to reflect the overall level of the signal. The calculation formula is: Where (n) is the number of data points, (x i ) is the value of each data point, and the formula is used to calculate the average of all values ​​in the data set; Variance calculation: Calculate the average of the square of the deviation between the signal value and the mean, which reflects the degree of fluctuation of the signal. The calculation formula of the variance is: Where (n) is the number of data points, (x i ) is the value of each data point, and the mean is the average of the data set. This formula averages the squares of the differences between each data point and the mean, providing a measure of the dispersion of the data distribution.

4. A chemical production control system according to claim 1, characterized in that: The fault diagnosis module is implemented in the following steps: (1) Data collection and preprocessing: Collect relevant chemical production process data and perform cleaning and normalization preprocessing operations; (2) Feature extraction: extracting meaningful features from the preprocessed data, which will be used for subsequent fault diagnosis; (3) Model training: Use the selected algorithm and collected data to train the fault diagnosis model; (4) Model evaluation and optimization: Evaluate the performance of the model and adjust parameters or optimize the model as needed; (5) Online fault diagnosis: Deploy the trained model to the actual system for real-time detection and diagnosis of faults.

5. A chemical production control system according to claim 4, characterized in that: The normalized calculation formula is as follows: Z-score normalization, also known as standardization, is a method of mapping data to a normal distribution with a mean of 0 and a standard deviation of 1. The specific calculation formula is as follows: Among them, x norm is the normalized data, (x) is the original data, (μ) is the mean of the original data, and (σ) is the standard deviation of the original data. This normalization method can eliminate the scale difference of the data so that different features or variables have the same scale.

6. A chemical production control system according to claim 4, characterized in that: The model training adopts a recurrent neural network (RNN) algorithm. The formula of RNN can be expressed as: ht = f(W·ht-1+U·xt+b)h t =f(W·h t-1 +U·x t +b)ht=f(W·ht-1+U·xt+b), where: ht represents the hidden state at the tth moment, W and U are weight matrices used to capture the correlation between the input and hidden state at different moments, b is the bias term, f is the activation function used to introduce nonlinear factors so that the model can capture complex patterns, and xt is the input at the tth moment. This formula reflects the recurrent connection in RNN, that is, the output at each moment depends not only on the input at the current moment, but also on the hidden state at the previous moment. This structure enables RNN to capture the temporal dependency and correlation in sequence data, making it suitable for processing time series, speech or text sequence data.

7. A chemical production control system according to claim 1, characterized in that: The data acquisition module collects various data in the chemical production process in real time; the data processing module processes and analyzes the collected data to extract key features related to the fault; The fault diagnosis module automatically diagnoses the fault based on the processed data; after the fault is diagnosed, the early warning module promptly issues an early warning signal so that the operator can take timely measures to deal with the fault.

8. A chemical production control system according to claim 1, characterized in that: The data acquisition module can collect temperature, pressure, flow, and liquid level parameters in real time through a variety of high-precision sensors, so as to collect various data in the chemical production process.

9. A chemical production control system according to claim 1, characterized in that: The fault diagnosis module has built-in rule-based diagnosis models and machine learning-based diagnosis models, and can automatically diagnose faults in the chemical production process based on processed data and extracted features.

10. A chemical production control system according to claim 1, characterized in that ,The early warning module has multiple early warning modes, including sound and light alarm, SMS notification, and ,email reminder, which can ensure the timeliness and effectiveness of the early warning ,information.

Citation Information

Cited By

  • Fault comprehensive diagnosis system and method suitable for aviation aerostatic press

    CN119801984A

  • Production control method and system for compound emulsifier for knitting oil agent

    CN120722857A

  • PVB wedge-shaped film production line fault diagnosis early warning system based on intelligent sensing

    CN121209444A