A data processing-based vibration table vibration anomaly alarm method

Abstract

The application discloses a kind of vibration table vibration abnormality alarm method based on data processing, is specifically related to vibration table vibration abnormality monitoring technical field.The method includes: step one, the each monitoring data of vibration table is collected and preprocessed, step two, the prediction model of each monitoring data of vibration table is constructed, step three, the each monitoring data of vibration table is obtained, step four, the vibration risk level of vibration table in prediction period is evaluated, step five, the response measure of vibration table vibration abnormality is executed.The application is by constructing the prediction model of each monitoring data of vibration table, to realize the prediction of each data of vibration table in prediction period, can pre-judge the risk of each parameter of vibration table monitoring, to make response measure in advance, reduce the risk of vibration table in working process.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of vibration table vibration anomaly monitoring, and particularly relates to a vibration table vibration anomaly alarm method based on data processing. BACKGROUND

[0002] The vibration table is used for simulating the vibration caused by the earthquake, wind or other power source, and is widely used in the experiment and test of the building structure, mechanical equipment, automobile industry, aerospace and other fields. As a kind of precision equipment, the running state of the vibration table is crucial to the experimental results. Through real-time monitoring and alarm, measures can be taken in time when the equipment fails or abnormity occurs, so as to avoid more serious damage or safety accidents. Therefore, it is necessary to study the vibration table vibration anomaly alarm technology.

[0003] The existing technology for monitoring the vibration anomaly of the vibration table meets the basic needs, but also has some defects, specifically: on the one hand, the vibration table vibration anomaly alarm technology is the key to guarantee the safety of the experiment and the stability of the equipment, but the current technology still has some deficiencies in detection accuracy, real-time performance, intelligence and multi-dimensional analysis, etc., it is difficult to find the vibration anomaly of the vibration table in time, causing the abnormal risk of the vibration table in the working process, and it is difficult to meet the production demand. On the other hand, the existing technology directly judges the data collected by the vibration table, but in the actual process, due to the large amount of data to be collected, it is difficult to make accurate judgment on the vibration anomaly of the vibration table in real time, the historical monitoring data is ignored, the prediction of the vibration monitoring data of the vibration table is not constructed, and it is difficult to realize the pre-discovery of the potential risk in the working process of the vibration table. SUMMARY

[0004] The purpose of the present application is to provide a vibration table vibration anomaly alarm method based on data processing, which solves the problems in the background art.

[0005] To solve the above technical problems, the technical scheme adopted by the present application is as follows: a vibration table vibration anomaly alarm method based on data processing, the method comprising: step one, collecting and preprocessing the monitoring data of the vibration table: collecting the monitoring data of each collection point of the vibration table in the target period, and normalizing the monitoring data of each collection point of the vibration table in the target period;

[0006] Step two, constructing a prediction model of the monitoring data of the vibration table: based on the historical monitoring data of each collection point of the vibration table in each period extracted from the database, constructing a prediction model of the monitoring data of the vibration table;

[0007] Step three, obtaining the predicted monitoring data of the vibration table: based on the normalized monitoring data of each collection point of the vibration table in the target period, the monitoring data of the vibration table is input into the prediction model of the monitoring data of the vibration table, and the predicted monitoring data of the vibration table in the prediction period is obtained.

[0008] Step four, evaluating the vibration risk level of the vibration table in the prediction period: based on the predicted monitoring data of the vibration table in the prediction period, the vibration risk coefficient of the vibration table in the prediction period is calculated, and the vibration risk level of the vibration table in the prediction period is evaluated.

[0009] Step five, executing the response measure of the vibration table vibration anomaly: implementing the response measure corresponding to the vibration risk level of the vibration table in the prediction period.

[0010] Compared with the prior art, the benefits of the present application are: first, the present application realizes the prediction of the monitoring data of the vibration table in the prediction period by constructing the prediction model of the monitoring data of the vibration table, which can make risk judgment on the monitoring parameters of the vibration table in advance, so as to make response measures in advance and reduce the risk of the vibration table in the working process.

[0011] Second, according to the predicted monitoring data of the vibration table in the prediction period, the vibration coefficient of the vibration table in the prediction period is analyzed and evaluated, so as to automatically generate the response measure of the vibration risk of the vibration table in the prediction period, which provides reliable technical scheme support for avoiding the vibration risk of the vibration table. BRIEF DESCRIPTION OF DRAWINGS

[0012] In order to more clearly illustrate the technical solutions in the embodiments of the present application or the prior art, the drawings needed to be used in the embodiments or the prior art description will be briefly introduced. Obviously, the drawings in the following description are only some embodiments of the present application, and other drawings can be obtained by those skilled in the art without creative labor.

[0013] Figure 1 The method flowchart of the present application.

[0014] Figure 2 The system structure connection diagram of the present application.

[0015] Figure 3 The network model diagram of the present application. DETAILED DESCRIPTION

[0016] The technical solutions in the embodiments of the present application will be clearly and completely described with reference to the drawings in the embodiments of the present application. Obviously, the described embodiments are only a part of the embodiments of the present application, rather than all the embodiments of the present application. Based on the embodiments in the present application, all other embodiments obtained by a person of ordinary skill in the art without creative effort belong to the scope of protection of the present application.

[0017] For the convenience of understanding the present application, some professional terms related to the present application are explained:

[0018] Vibration table: a vibration table is used to simulate the vibration caused by earthquakes, wind or other power sources, and is widely used in experiments and tests in the fields of building structures, mechanical equipment, automobile industry, aerospace, etc.

[0019] Normalization: normalization is to scale data to a fixed range. Normalization is very important to ensure that different features contribute fairly to the model, especially when using gradient descent, which can speed up the training process.

[0020] LSTM: long short-term memory neural network is an effective method for constructing a prediction model for vibration table vibration monitoring data. LSTM is a special recurrent neural network (RNN) that is good at processing and predicting time-based sequence data.

[0021] Sliding window: the core idea of the sliding window method is to slide a fixed-size window over the time series data, extracting sub-sequences from each window. These sub-sequences will be used as input to the model. Each time the window slides, the data moves by a fixed step.

[0022] Gradient descent: by calculating the gradient of the objective function at the current parameter point (i.e. the derivative of the objective function), then updating the model parameters in the opposite direction of the gradient (i.e. the negative gradient direction), so as to gradually reduce the value of the objective function.

[0023] Embodiment one:

[0024] Referring to Figure 1 The present application provides a vibration table vibration anomaly alarm method based on data processing, which comprises the following steps: step one, collecting and preprocessing the monitoring data of the vibration table: collecting the monitoring data of each collection point of the vibration table in the target period, and normalizing the monitoring data of each collection point of the vibration table in the target period.

[0025] In the specific embodiments of the present application, the method for collecting the monitoring data of each collection point of the vibration table in the target period is as follows: the monitoring data of the vibration table includes vibration acceleration, vibration frequency, vibration displacement, working temperature and working voltage.

[0026] The acceleration sensor installed on the vibration table collects the vibration frequency at a collection interval The vibration acceleration of the vibration table is collected as a collection interval The collection frequency extracted from the database is represented.

[0027] The velocity sensor installed on the vibration table collects the vibration frequency at a collection interval The vibration frequency of the vibration table is collected as a collection interval.

[0028] The displacement sensor installed on the vibration table collects the vibration frequency at a collection interval The vibration displacement of the vibration table is collected as a collection interval.

[0029] The temperature sensor installed on the vibration table collects the vibration frequency at a collection interval The working temperature of the vibration table is collected as a collection interval.

[0030] The voltage sensor installed on the vibration table collects the vibration frequency at a collection interval The working voltage of the vibration table is collected as a collection interval.

[0031] It should be noted that in the specific embodiments of the present application, the collection frequency extracted from the database is specifically set as follows: the vibration risk level of the vibration table in the previous period of the target period is extracted from the database, if the vibration risk level of the vibration table in the previous period of the target period is primary, the collection frequency in the target period is set to , if the vibration risk level of the vibration table in the previous period of the target period is intermediate, the collection frequency in the target period is set to , if the vibration risk level of the vibration table in the previous period of the target period is high, the collection frequency in the target period is set to , wherein is stored in the database and satisfies , for example .

[0032] In the specific embodiments of the present application, the monitoring data of each collection point of the vibration table in the target period is normalized, and the specific analysis method is as follows: the monitoring data of each collection point of the vibration table in the target period is recorded as , wherein , represents the number of each collection point, represents the total number of collection points, , represents the number of each monitoring data, represents the total number of monitoring data.

[0033] Extract the maximum value and minimum value of each monitoring data in the target period from each monitoring data of each collection point of the vibration table in the target period, and mark them as and .

[0034] The normalized monitoring data of each collection point of the vibration table in the target period is calculated by the formula: .

[0035] It should be noted that in the specific embodiments of the present application, the calculation formula of the normalized monitoring data of each collection point of the vibration table in the target period is based on the data normalization processing method, which realizes the scaling of the monitoring data of each collection point of the vibration table in the target period to a fixed range , and normalization is very important to ensure the fair contribution of different features to the model.

[0036] Step two, constructing a prediction model of each monitoring data of the vibration table: based on the historical monitoring data of each collection point of the vibration table in each period extracted from the database, a prediction model of each monitoring data of the vibration table is constructed.

[0037] In specific embodiments of the present application, the specific analysis method of the prediction model of each monitoring data of the vibration table is: the detailed method steps of constructing the prediction model of each monitoring data of the vibration table by LSTM long short-term memory neural network are: A1: extracting window size and step from the database, creating a sliding window, and segmenting the normalized historical monitoring data of each collection point of the vibration table in each period extracted from the database to obtain the subsequence of the normalized historical monitoring data of each collection point of the vibration table in each period.

[0038] It should be noted that in the specific embodiments of the present application, the specific analysis method of the subsequence of the normalized historical monitoring data of each collection point of the vibration table in each period is: for example, the vibration frequency data of each collection point of the vibration table in the historical first period is recorded as , the window size is set to , the step is set to , and the subsequence is extracted by the sliding window, then the first window is obtained: , the second window is: , the third window is: , and the fourth window is: .

[0039] ​It should be noted that in the specific embodiments of the present application, the sliding window is a powerful method that can effectively process time series data. In building the prediction model of each monitoring data of the vibration table, the sliding window technology can help the model capture the local characteristics of the time series data and provide effective support for subsequent machine learning tasks.

[0040] A2: The subsequence of the normalized historical monitoring data of each collection point of the vibration table in each period is taken as the training input of the LSTM long short-term memory neural network.

[0041] It should be noted that in the specific embodiments of the present application, the LSTM long short-term memory neural network specifically includes an input layer, an LSTM layer, a fully connected layer, and an output layer.

[0042] The output layer is the first layer of the neural network, which receives data input from the outside. In time series problems, the input layer receives feature data at each time. The input layer passes these data to the LSTM layer as initial information for the model.

[0043] The role of the LSTM layer is to capture the long and short-term dependencies in the time series. It can effectively remember and forget past information, thereby solving the gradient vanishing or explosion problem that standard RNNs may encounter in long sequences.

[0044] The fully connected layer is responsible for mapping the features extracted by the LSTM layer to higher-level representations, usually used for classification or regression tasks. Through the fully connected layer, the network can learn the mapping from the features extracted from the time series data to the prediction values of the monitoring data of the vibration table.

[0045] The role of the output layer is to generate the final prediction or classification result according to the processing results of the previous layers.

[0046] A3: The loss function of the LSTM long short-term memory neural network is constructed by the formula: wherein represents the true value of the training input of the LSTM long short-term memory neural network, represents the prediction value of the output of the LSTM long short-term memory neural network.

[0047] A4: The gradient descent method is used to update the weights of the LSTM long short-term memory neural network by the formula: represents the weights of the LSTM long short-term memory neural network, represents the learning rate of the LSTM long short-term memory neural network extracted from the database.

[0048] ​​A5: the bias parameters of the LSTM long short-term memory neural network are updated by the formula: , wherein represents the bias parameters of the LSTM long short-term memory neural network.

[0049] The present application realizes the prediction of the monitoring data of the vibration table in the prediction period by constructing the prediction model of each monitoring data of the vibration table, can make risk judgment on the monitoring parameters of the vibration table in advance, and thus makes response measures in advance to reduce the risk of the vibration table in the working process.

[0050] Step three, obtaining the predicted monitoring data of the vibration table: based on the normalized monitoring data of each collection point of the vibration table in the target period, the normalized monitoring data of each collection point of the vibration table in the target period is input into the prediction model of the monitoring data of the vibration table, and the predicted monitoring data of the vibration table in the prediction period is obtained.

[0051] In the specific embodiments of the present application, the specific analysis method for obtaining the predicted monitoring data of the vibration table in the prediction period is as follows: based on the normalized monitoring data of each collection point of the vibration table in the target period, the subsequence of the normalized monitoring data of each collection point of the vibration table in the target period is obtained after sliding window processing, which is used as the input of the prediction model of the monitoring data of the vibration table, and the predicted monitoring data of the vibration table in the prediction period is obtained, denoted as .

[0052] Step four, evaluating the vibration risk level of the vibration table in the prediction period: based on the predicted monitoring data of the vibration table in the prediction period, the vibration risk coefficient of the vibration table in the prediction period is calculated, and the vibration risk level of the vibration table in the prediction period is evaluated.

[0053] In the specific embodiments of the present application, the specific analysis method for calculating the vibration risk coefficient of the vibration table in the prediction period is as follows: , the abnormal judgment value of the predicted monitoring data of the vibration table in the prediction period is calculated by the formula: , wherein represents the normal working interval of the monitoring data of the vibration table stored in the database.

[0054] ​It should be noted that in the specific embodiments of the present application, the calculation method of the abnormal judgment value of each monitoring data of the vibration table predicted in the prediction period is specifically: for example, if the predicted vibration frequency of the vibration table in the prediction period is not in the normal working interval of the vibration frequency of the vibration table, the abnormal judgment value of the predicted vibration frequency of the vibration table in the prediction period is set to 1, and if the predicted vibration frequency of the vibration table in the prediction period is in the normal working interval of the vibration frequency of the vibration table, the abnormal judgment value of the predicted vibration frequency of the vibration table in the prediction period is set to 0.

[0055] It should be noted that in the specific embodiments of the present application, the abnormal judgment value of each monitoring data of the vibration table predicted in the prediction period reflects whether each monitoring data of the vibration table predicted in the prediction period is in the normal working state interval.

[0056] It should be noted that in the specific embodiments of the present application, the normal working interval of each monitoring data of the vibration table is artificially set by technicians according to the actual working each monitoring data of the vibration table and the working state of the vibration table after comprehensive analysis, and is stored in the database.

[0057] The vibration risk coefficient of the vibration table in the prediction period is calculated by the formula: , wherein represents a natural constant.

[0058] It should be noted that in the specific embodiments of the present application, the calculation method of the vibration risk coefficient of the vibration table in the prediction period comprehensively considers the abnormal judgment value of each monitoring data of the vibration table predicted in the prediction period, and the greater the sum of the abnormal judgment values of each monitoring data of the vibration table predicted in the prediction period, the greater the vibration risk of the vibration table in the prediction period is analyzed, and the vibration risk coefficient of the vibration table in the prediction period is used as a reference value reflecting the vibration risk of the vibration table in the prediction period.

[0059] In the specific embodiments of the present application, the specific analysis method of the vibration risk level of the vibration table in the prediction period is based on the vibration risk coefficient of the vibration table in the prediction period , if is in the first vibration risk interval extracted from the database, it is determined that the vibration risk level of the vibration table in the prediction period is primary.

[0060] If is in the second vibration risk interval extracted from the database, it is determined that the vibration risk level of the vibration table in the prediction period is intermediate.

[0061] If ​If the vibration risk third interval is extracted from the database, it is determined that the vibration table vibration risk level in the prediction period is high.

[0062] It should be noted that in the specific embodiments of the present application, the vibration risk first interval, the vibration risk second interval and the vibration risk third interval are set artificially by technicians based on the vibration risk coefficients of the vibration table in each period and the actual working state of the subsequent vibration table, and are stored in the database. For example, the vibration risk first interval is: , the vibration risk first interval is: , the vibration risk first interval is: .

[0063] Step five, executing the vibration table vibration abnormality response measure: implementing the response measure corresponding to the vibration table vibration risk level in the prediction period.

[0064] In the specific embodiments of the present application, the specific analysis method of implementing the response measure corresponding to the vibration table vibration risk level in the prediction period is: based on the abnormality judgment value of each monitoring data of the vibration table in the prediction period, extracting the abnormal monitoring parameters of the vibration table in the prediction period, extracting the abnormal processing method of the abnormal monitoring parameters of the historical vibration table from the database, generating the vibration risk response scheme of the vibration table in the prediction period, issuing the vibration risk response scheme of the vibration table in the prediction period to the vibration table, and implementing the vibration risk response measure of the vibration table in the prediction period.

[0065] It should be noted that in the specific embodiments of the present application, the specific method steps of implementing the vibration risk response measure of the vibration table in the prediction period include the following information: the abnormal monitoring parameters of the vibration table in the prediction period, the response measure of the abnormal monitoring parameters of the vibration table in the prediction period and the vibration risk response emergency identifier of the vibration table in the prediction period. The vibration risk response emergency identifier of the vibration table in the prediction period is extracted from the vibration risk response measure of the vibration table in the prediction period. If the vibration risk response emergency identifier of the vibration table in the prediction period is , the vibration table is immediately stopped working, the response measure of the abnormal monitoring parameters of the vibration table in the prediction period is extracted, and the vibration risk response of the vibration table is performed.

[0066] If the vibration risk response emergency identifier of the vibration table in the prediction period is , the vibration table continues to work for a period of time, the response measure of the abnormal monitoring parameters of the vibration table in the prediction period is extracted, and the vibration risk response of the vibration table is performed.

[0067] If the vibration table continues to operate, the response measures for each parameter of the vibration table abnormal monitoring within the prediction period are extracted, and the vibration risk response of the vibration table is carried out.

[0068] It should be noted that, in a specific embodiment of the present invention, the working state of the vibration table during the vibration risk response process is controlled according to the emergency identifier of the vibration table vibration risk response within the prediction period. Since the vibration risk of the vibration table within the prediction period is predicted based on the monitoring parameters of the vibration table within the target period, the working efficiency of the vibration table can be maintained during the pre-response vibration risk response process.

[0069] In a specific embodiment of the present invention, the specific analysis method for extracting the abnormal monitoring parameters of the vibration table within the prediction period is as follows: based on the abnormal judgment values ​​of each monitoring data predicted by the vibration table within the prediction period, if... If the value is 1, then the predicted value of the vibration table within the prediction period will be the first predicted value. Each monitoring data point is recorded as an anomaly monitoring parameter.

[0070] According to the prediction of the vibration table within the prediction period, the first The method of analyzing monitoring data is used to analyze the abnormal judgment values ​​of each monitoring data predicted by the vibration table within the prediction period, and to obtain the abnormal monitoring parameters of the vibration table within the prediction period.

[0071] In a specific embodiment of the present invention, the specific analysis method for generating the vibration risk response scheme of the shaking table within the prediction period is as follows: based on the abnormal monitoring parameters of the shaking table within the prediction period, the response measures of the abnormal monitoring parameters of the historical shaking table are extracted from the database to generate the vibration risk response scheme of the shaking table within the prediction period.

[0072] Based on the vibration risk level of the shaking table within the prediction period, an emergency identifier for the vibration risk response of the shaking table within the prediction period is generated.

[0073] It should be noted that, in a specific embodiment of the present invention, the specific analysis method for generating the vibration risk response emergency identifier for the vibration table within the prediction period is as follows: for example, if the vibration risk level of the vibration table within the prediction period is primary, then the vibration risk response emergency identifier for the vibration table within the prediction period is set to... If the vibration risk level of the shaking table within the prediction period is medium, then the emergency identifier for the vibration risk response of the shaking table within the prediction period will be set to... If the vibration risk level of the shaking table during the prediction period is high, then the emergency identifier for the vibration risk response of the shaking table during the prediction period will be set to [high level]. .

[0074] The vibration table vibration risk scheme in the prediction period includes: vibration table abnormality monitoring parameters in the prediction period, response measures for vibration table abnormality monitoring parameters in the prediction period, and vibration table vibration risk response emergency identifiers in the prediction period.

[0075] According to the monitoring data predicted by the vibration table in the prediction period, the vibration table vibration coefficient in the prediction period is analyzed and evaluated, so that the response measures of the vibration table vibration risk in the prediction period are automatically generated, and reliable technical scheme support is provided for avoiding the vibration table vibration risk.

[0076] Embodiment two:

[0077] It should be noted that, in the specific embodiments of the present application, referring to Figure 2 The present application provides a system for executing a vibration table vibration abnormality alarm method based on data processing, which specifically includes: a monitoring data acquisition and preprocessing module, a prediction model construction module, a predicted monitoring data acquisition module, a vibration risk level evaluation module, a vibration abnormality response module, and a database.

[0078] The monitoring data acquisition and preprocessing module is used to acquire each monitoring data of each collection point of the vibration table in the target period and perform normalization processing.

[0079] The prediction model construction module is used to construct a prediction model of each monitoring data of the vibration table through an LSTM long short-term memory neural network.

[0080] The predicted monitoring data acquisition module is used to acquire each monitoring data predicted by the vibration table in the prediction period according to each monitoring data of each collection point of the vibration table in the target period after normalization processing.

[0081] The vibration risk level evaluation module is used to calculate the vibration risk coefficient of the vibration table in the prediction period according to each monitoring data predicted by the vibration table in the prediction period, and evaluate the vibration risk level of the vibration table in the prediction period.

[0082] The vibration abnormality response module is used to implement the response measures corresponding to the vibration risk level of the vibration table in the prediction period.

[0083] The database is used to store the collection frequency of each monitoring data of the vibration table, the window size and step size of the sliding window, the normal working interval of each monitoring data of the vibration table, the normalized historical monitoring data of each collection point of the vibration table in each period, and the vibration risk first interval, the vibration risk second interval, and the vibration risk third interval.

[0084] Embodiment three:

[0085] It should be noted that in the specific embodiments of the present application, referring to Figure 3 The network model of the vibration table vibration anomaly alarm method based on data processing is specifically implemented, which comprises a vibration table, a monitoring data uploading end, a cloud platform and a response measure issuing end.

[0086] The vibration table collects the monitoring data of the vibration table through various sensors distributed on the vibration table, receives the vibration table vibration risk response scheme generated by the cloud platform within the prediction period, and performs alarm and response processing.

[0087] The monitoring data uploading end collects the monitoring data collected by various sensors distributed on the vibration table through a network transmission module, and sends the monitoring data to the cloud platform.

[0088] The cloud platform is used to receive the monitoring data of the vibration table, and generate the monitoring data of the vibration table within the prediction period through the prediction model of the monitoring data of the vibration table, analyze and evaluate, if the vibration table has vibration risk, and generate the vibration table vibration risk response scheme within the prediction period.

[0089] The response measure issuing end issues the generated vibration table vibration risk response scheme within the prediction period to the vibration table through a network transmission module.

[0090] The above content is only an example and description of the concept of the present application, and those skilled in the art can make various modifications or supplements or use similar ways to replace the described specific embodiments, as long as they do not deviate from the concept of the present application or exceed the scope defined by the present application, which shall belong to the protection scope of the present application.

Claims

1. A data processing-based vibration table vibration anomaly alarm method, characterized in that, The method comprises: Step one, collecting and preprocessing the monitoring data of the vibration table: collecting the monitoring data of each collection point of the vibration table in the target period, and normalizing the monitoring data of each collection point of the vibration table in the target period; Step two, constructing a prediction model of the monitoring data of the vibration table: based on the historical monitoring data of each collection point of the vibration table in each period extracted from the database, a prediction model of the monitoring data of the vibration table is constructed; Step three, obtaining the predicted monitoring data of the vibration table: based on the normalized monitoring data of each collection point of the vibration table in the target period, the normalized monitoring data is input into the prediction model of the monitoring data of the vibration table, and the predicted monitoring data of the vibration table in the prediction period is obtained; Step four, evaluating the vibration risk level of the vibration table in the prediction period: based on the predicted monitoring data of the vibration table in the prediction period, the vibration risk coefficient of the vibration table in the prediction period is calculated, and the vibration risk level of the vibration table in the prediction period is evaluated; Step five, executing the response measures for the vibration anomaly of the vibration table: the response measures corresponding to the vibration risk level of the vibration table in the prediction period are implemented; The specific analysis method for constructing the prediction model of the monitoring data of the vibration table is: The detailed method steps for constructing the prediction model of the monitoring data of the vibration table by LSTM long short-term memory neural network are: A1: extracting window size from database and step size , creating a sliding window, dividing the normalized historical monitoring data of each collection point of the vibration table in each period to obtain the subsequence of the normalized historical monitoring data of each collection point of the vibration table in each period; A2: The subsequence of the normalized historical monitoring data of each collection point of the vibration table in each period is used as the training input of the LSTM long short-term memory neural network; A3: constructing a loss function of the LSTM long short-term memory neural network by a formula: , wherein , wherein represents a true value of a training input of the LSTM long short-term memory neural network, represents a predicted value of an output of the LSTM long short-term memory neural network; A4: the gradient descent method is used to update the weight of the LSTM long short-term memory neural network through the formula: , wherein represents the weight of the LSTM long short-term memory neural network, represents the learning rate of the LSTM long short-term memory neural network extracted from the database, represents a partial derivative calculation identifier; A5: the bias parameters of the LSTM long short-term memory neural network are updated by the formula: , wherein represents the bias parameters of the LSTM long short-term memory neural network. The specific analysis method for calculating the vibration risk coefficient of the vibration table in the prediction period is: Based on the monitoring data predicted by the shaking table within the prediction period Through the formula: The anomaly judgment values ​​of each monitoring data predicted by the vibration table within the prediction period are calculated. ,in This indicates the normal operating range of each monitoring data point from the vibration table stored in the database; The vibration table vibration risk coefficient in the prediction period is calculated by the formula: , wherein , wherein represents a natural constant; The specific analysis method for evaluating the vibration risk level of the vibration table in the prediction period is: Based on a vibration table vibration risk coefficient in a prediction period , if a vibration risk first interval extracted from a database, a vibration table vibration risk level in a prediction period is determined as primary; If If the vibration risk second interval extracted from the database is reached, it is determined that the vibration table vibration risk level in the prediction period is moderate. If If the vibration risk is in the third interval extracted from the database, it is determined that the vibration table vibration risk level in the prediction period is high. The specific analysis method for generating the vibration risk response scheme of the vibration table in the prediction period is: Based on the abnormal monitoring parameters of the vibration table in the prediction period, the response measures for the abnormal monitoring parameters of the vibration table in the prediction period are extracted from the database, and the vibration risk response scheme of the vibration table in the prediction period is generated; Based on the vibration risk level of the vibration table in the prediction period, the vibration risk response emergency identifier in the prediction period is generated; The vibration risk scheme of the vibration table in the prediction period comprises: the abnormal monitoring parameters of the vibration table in the prediction period, the response measures for the abnormal monitoring parameters of the vibration table in the prediction period, and the vibration risk response emergency identifier of the vibration table in the prediction period.

2. The vibration table vibration anomaly alarm method based on data processing according to claim 1, characterized in that, The specific method for collecting the monitoring data of each collection point of the vibration table in the target period is: The monitoring data of the vibration table comprises: vibration acceleration, vibration frequency, vibration displacement, working temperature and working voltage; The collection frequency is collected by an acceleration sensor installed on the vibration table The vibration acceleration of the vibration table is collected as a collection interval The collection frequency is collected from the database The frequency is collected by the speed sensor installed on the vibration table The vibration frequency of the vibration table is collected as the collection interval; The collection of the vibration displacement of the vibration table is carried out as a collection interval The collection of the vibration displacement of the vibration table is carried out as a collection interval The acquisition frequency is set by the temperature sensor installed on the vibration table The acquisition of the working temperature of the vibration table is performed as the acquisition interval; The frequency is collected by the voltage sensor installed on the vibration table The working voltage of the vibration table is collected as the collection interval.

3. The vibration table vibration anomaly alarm method based on data processing according to claim 2, characterized in that, The specific analysis method for normalizing the monitoring data of each collection point of the vibration table in the target period is: The monitoring data of each collection point of the vibration table in the target period is recorded as , wherein , represents the number of each collection point, represents the total number of collection points, , represents the number of each monitoring data, represents the total number of monitoring data; The maximum value and the minimum value of each monitoring data in the target period are extracted from each monitoring data of each collection point of the vibration table in the target period, and are respectively denoted as and ; The normalized monitoring data of each collection point of the vibration table in the target period is calculated by the formula: .​ 4. The vibration table vibration anomaly alarm method based on data processing according to claim 3, characterized in that, The specific analysis method for obtaining the predicted monitoring data of the vibration table in the prediction period is: Based on the normalized monitoring data of each acquisition point of the vibration table in the target period, the subsequence of the normalized monitoring data of each acquisition point of the vibration table in the target period is obtained after sliding window processing, as the input of the prediction model of the monitoring data of the vibration table, and the predicted monitoring data of the vibration table in the prediction period is obtained, denoted as .

5. The vibration anomaly alarm method of vibration table based on data processing according to claim 4, characterized in that, The specific analysis method for implementing the response measures corresponding to the vibration risk level of the vibration table in the prediction period is: Based on the abnormality judgment value of each monitoring data of the vibration table in the prediction period, the abnormal monitoring parameters of the vibration table in the prediction period are extracted, the abnormality processing method of the abnormal monitoring parameters of the historical vibration table is extracted from the database, the vibration risk response scheme of the vibration table in the prediction period is generated, the vibration risk response scheme of the vibration table in the prediction period is issued to the vibration table, and the vibration risk response measure of the vibration table in the prediction period is implemented.

6. The vibration table vibration anomaly alarm method based on data processing according to claim 5, characterized in that, The specific analysis method of the abnormal monitoring parameters of the vibration table in the prediction period is that: Based on the abnormality judgment value of each monitoring data predicted by the vibration table in the prediction period, if is 1, the predicted first monitoring data of the vibration table in the prediction period is recorded as an abnormal monitoring parameter; According to the prediction of the vibration table within the prediction period, the first The analysis method of monitoring data is used to analyze the abnormal judgment values ​​of each monitoring data predicted by the vibration table within the prediction period, and to obtain the abnormal monitoring parameters of the vibration table within the prediction period.

Images