Real-time monitoring method and system for ship construction engineering

Abstract

The invention discloses a real-time monitoring method and system for ship construction engineering, and relates to the field of ship construction engineering. The method comprises the following steps: acquiring operation state parameters of to-be-monitored processing equipment in a processing process at a sampling frequency f, constructing a first time sequence data sequence of an ith operation state parameter, extracting operation state parameter node data in adjacent time periods, and performing cross validation by using KPSS (Knowledge Pattern Status) inspection and ADF (Advanced Development Function) inspection to obtain a first time sequence data sequence of the ith operation state parameter; classifying the change state of the node data into a stable change state or a non-stable change state, constructing a first prediction model when the node data is in the stable change state and a second prediction model when the node data is in the non-stable change state, and obtaining running state parameter node prediction data in adjacent time periods, and constructing a second time sequence data sequence for driving the digital twinborn model of the ship construction engineering, thereby realizing virtual simulation and real-time monitoring of the processing process of the processing equipment to be monitored.

Description

Technical Field

[0001] The present invention relates to the field of shipbuilding engineering, and in particular to a real-time monitoring method and system for shipbuilding engineering. Background Art

[0002] Shipbuilding is a complex and time-consuming engineering process involving multiple stages and numerous processes, including material processing, component assembly, welding, and painting. Traditionally, shipbuilding monitoring relies primarily on manual inspections and periodic quality checks. This is not only inefficient but also difficult to achieve real-time monitoring of the entire construction process. With the advancement of digital technology, digital twin technology is gradually being applied to shipbuilding. By establishing a real-time mapping between virtual models and physical entities, comprehensive monitoring and optimization of the shipbuilding process can be achieved.

[0003] Real-time parameter monitoring during the processing is one of the important projects in shipbuilding process monitoring. In the existing technology, by installing multiple sensors on the processing equipment, multi-source operating status parameters are collected in real time, and the equipment's operating parameters and sensor data are integrated into the digital twin model. Using the digital twin model, real-time monitoring of the operating status of processing equipment including cutting equipment and welding equipment is achieved, thereby identifying abnormal conditions and potential faults in equipment operation.

[0004] However, the existing monitoring process described above has certain limitations and contradictions. These are primarily manifested in the following ways: First, building a digital twin model requires collecting massive amounts of real-time operating status parameters from multiple sources. The greater the data density, the more accurate the real-time mapping between the constructed virtual model and the physical entity. Second, a higher frequency of data collection increases not only the amount of real-time data to be processed but also the difficulty of data calculation, which in turn reduces the timeliness of model construction. Furthermore, the durability and reliability requirements for sensor equipment also increase dramatically. Therefore, balancing the contradiction between model accuracy and timeliness caused by real-time parameter collection during the manufacturing process is a prerequisite that requires careful consideration when building a digital twin model and a key step in achieving real-time monitoring of the operating status of manufacturing equipment. To this end, a real-time monitoring method and system for shipbuilding engineering is proposed. Summary of the Invention

[0005] The main purpose of the present invention is to provide a real-time monitoring method and system for shipbuilding projects, which can effectively solve the problems in the background technology.

[0006] In order to achieve the above purpose, the technical solution adopted by the present invention is:

[0007] A real-time monitoring method for shipbuilding projects, comprising:

[0008] Step 1: Collect the operating status parameters of the monitored processing equipment during the processing at a sampling frequency of f, and construct the first time series data sequence Rs of the i-th operating status parameter i (t) = {rs i1 ,rs i2 ,...,rs in}, where t represents time; rs in is the sampling value of the operating status parameter when t=n;

[0009] Step 2: Extract the first time series data sequence Rs i (t) The operating status parameter node data {rs i1 ,rs i2 ,...,rs iq}, using KPSS test and ADF test for cross validation, the change state of the extracted node data is determined, and the change state of the node data is classified as a stable change state or a non-stationary change state;

[0010] Step 3: Based on the state classification result of the node data, a prediction model is constructed using the extracted node data, wherein the prediction model includes a first prediction model when the node data is in a stable change state and a second prediction model when the node data is in a non-stationary change state;

[0011] Step 4: According to the constructed prediction model, obtain the node prediction data of the operating state parameter in the period from t=q to t=q+1, and add the obtained node prediction data to the first time series data sequence Rs i (t) forms the second time series data sequence Rs' i (t);

[0012] Step 5: Use the acquired second time series data sequence Rs' i (t) Drives the creation of a digital twin model for shipbuilding projects, through which the processing of the monitored processing equipment is virtually simulated and monitored in real time.

[0013] A real-time monitoring system for shipbuilding projects, comprising:

[0014] A state parameter acquisition module is used to collect the operating state parameters of the processing equipment to be monitored during the processing at a sampling frequency of f, and to construct a first time series data sequence of the i-th operating state parameter;

[0015] Data verification module, used to extract the first time series data sequence Rs iThe node data of the operating status parameters in the period from t = 1 to t = q in (t) are cross-validated using the KPSS test and the ADF test to determine the change state of the extracted node data and classify the change state of the node data as a steady change state or a non-steady change state;

[0016] A prediction model building module is used to build a prediction model using the extracted node data according to the state classification result of the node data, including a first prediction model when the node data is in a stable change state and a second prediction model when the node data is in a non-stationary change state;

[0017] The data reconstruction module is used to obtain the node prediction data of the operating state parameter in the period from t=q to t=q+1 according to the constructed prediction model, and add the obtained node prediction data to the first time series data sequence Rs i (t) forms the second time series data sequence Rs' i (t);

[0018] A digital twin model construction module is used to construct a digital twin model for virtual simulation and real-time monitoring of the processing process of the monitored processing equipment;

[0019] Model driving module, used to obtain the second time series data sequence Rs' i (t) Importing it into the digital twin model, thereby driving the created digital twin model of the shipbuilding project to perform virtual simulation and real-time monitoring on the processing process of the monitored processing equipment.

[0020] The system further includes a memory, a processor, and a computer program stored in the memory and executable on the processor.

[0021] Furthermore, the processing equipment includes cutting equipment and welding equipment, wherein:

[0022] The operating state parameters of the cutting equipment include one or more combinations of cutting power, cutting speed, cutting air pressure and cutting head height;

[0023] The operating status parameters of the welding equipment include one or more combinations of welding current, welding voltage, welding speed, wire consumption, gas flow, energy consumption value and welding quality parameters.

[0024] Furthermore, the specific process of step 2 includes the following steps:

[0025] Step S21: Explicitly test the hypothesis

[0026] For the ADF test:

[0027] Set the null hypothesis H0 as: the extracted node data sequence has a unit root, that is, a non-stationary change state;

[0028] The alternative hypothesis H1 is: there is no unit root in the extracted node data sequence, that is, a steady change state;

[0029] For KPSS test:

[0030] Set the null hypothesis H0 as: the extracted node data sequence is in a stable changing state;

[0031] The alternative hypothesis H1 is: the extracted node data sequence is in a non-stationary changing state;

[0032] Step S22: Perform ADF test and obtain the test result of ADF test

[0033] Step S23: Perform KPSS test and obtain the test result of KPSS test

[0034] Step S24: Cross-validation results

[0035] When the results of the ADF test and the KPSS test are consistent

[0036] If both indicate that the extracted node data sequence is in a steady-changing state, then the extracted node data sequence is considered to be in a steady-changing state;

[0037] If both indicate that the extracted node data sequence is in a non-stationary change state, then the extracted node data sequence is considered to be in a non-stationary change state;

[0038] When the results of the ADF test and the KPSS test are inconsistent

[0039] Perform first-order or higher-order difference processing on the extracted node data series until the results of the ADF test and the KPSS test are consistent.

[0040] Furthermore, the expression of the first prediction model is:

[0041]

[0042] Where, is the node prediction data of the operating state parameter obtained by the first prediction model in the period from t = q to t = q + 1; c is the constant coefficient; ε is the error coefficient; μ k is the model coefficient; p is an integer between 1 and q-1.

[0043] Furthermore, the expression of the second prediction model is:

[0044]

[0045] Where, is the node prediction data of the operating state parameter obtained by the second prediction model in the period from t = q to t = q + 1; c is the constant coefficient; ε is the error coefficient; μ k is the model coefficient; p is an integer between 1 and q-1; ε (q-k+1) is the random error term; ν k is the random error coefficient.

[0046] Furthermore, the ADF test process is:

[0047] Calculate the ADF test statistic;

[0048] If the calculated p-value is less than the set significance level and the ADF test statistic is less than the set critical value, the null hypothesis is rejected and it is considered that the extracted node data sequence is in a stable change state;

[0049] If the calculated p-value is greater than the set significance level, or the ADF test statistic is greater than the set critical value, the null hypothesis is accepted and it is considered that the extracted node data sequence is in a non-stationary change state;

[0050] The KPSS inspection process is as follows:

[0051] Calculate the KPSS test statistic;

[0052] If the calculated KPSS test statistic is less than the set critical value and the calculated p-value is greater than the significance level, the null hypothesis cannot be rejected and it is considered that the extracted node data sequence is in a stable change state;

[0053] If the calculated KPSS test statistic is greater than the set critical value, or the calculated p-value is less than the significance level, the null hypothesis of stationarity is rejected, and it is considered that the extracted node data sequence is in a non-stationary change state.

[0054] Furthermore, the significance level is set at 0.05.

[0055] The present invention has the following beneficial effects:

[0056] Compared with the existing technology, the operating status parameters of the processing equipment to be monitored during the processing are collected at a sampling frequency of f, the first time series data sequence of the i-th operating status parameter is constructed, the operating status parameter node data in the time period from t = 1 to t = q in the first time series data sequence is extracted, and the KPSS test and ADF test are used for cross-validation to classify the change state of the node data as a steady change state or a non-steady change state. A first prediction model is constructed when the node data is in a steady change state and a second prediction model is constructed when the node data is in a non-steady change state, and the node prediction of the operating status parameter in the time period from t = q to t = q+1 is obtained. Data, the acquired node prediction data is added to the first time series data sequence to form a second time series data sequence, and the digital twin model of the shipbuilding project is driven by the acquired second time series data sequence to realize virtual simulation and real-time monitoring of the processing process of the monitored processing equipment. By using this solution, the amount of data of the collected operating status parameters can be expanded according to the model construction requirements on the basis of keeping the data acquisition frequency unchanged, which is conducive to improving the accuracy of the real-time mapping between the virtual model and the physical entity, and at the same time can reduce the durability and reliability requirements of the sensing equipment, and improve the timeliness of the digital twin model construction and monitoring process. BRIEF DESCRIPTION OF THE DRAWINGS

[0057] Figure 1 This is a flow chart of a real-time monitoring method for shipbuilding projects according to the present invention;

[0058] Figure 2 The figure is a structural diagram of a real-time monitoring system for shipbuilding projects according to the present invention. DETAILED DESCRIPTION

[0059] The present invention will be further described below in conjunction with specific embodiments. The accompanying drawings are for illustrative purposes only and represent only schematic diagrams rather than actual drawings. They should not be understood as limiting the present invention. In order to better illustrate the specific embodiments of the present invention, some parts of the drawings may be omitted, enlarged or reduced, and do not represent the size of the actual product.

[0060] The specific implementation process of the technical solution of the present invention includes the following steps:

[0061] Step 1: Collect the operating status parameters of the processing equipment to be monitored during the processing at a sampling frequency of f.

[0062] Among them, the processing equipment includes cutting equipment and welding equipment. The operating status parameters of the cutting equipment include one or more combinations of cutting power, cutting speed, cutting gas pressure and cutting head height; the operating status parameters of the welding equipment include one or more combinations of welding current, welding voltage, welding speed, wire consumption, gas flow, energy consumption value and welding quality parameters.

[0063] Step 2: Construct the first time series data sequence Rs of the i-th operating status parameter i (t) = {rs i1 ,rs i2 ,...,rs in}, where t represents time; rs in is the sampling value of the operating status parameter when t=n;

[0064] Step 3: Extract the first time series data sequence Rs i (t) The operating status parameter node data {rs i1 ,rs i2 ,...,rs iq}, using KPSS test and ADF test for cross validation, the change state of the extracted node data is determined, and the change state of the node data is classified as a stable change state or a non-stationary change state;

[0065] The specific process of cross-validation includes the following steps:

[0066] Step S31: Clearly test the hypothesis

[0067] For the ADF test:

[0068] Set the null hypothesis H0 as: the extracted node data sequence has a unit root, that is, a non-stationary change state;

[0069] The alternative hypothesis H1 is: there is no unit root in the extracted node data sequence, that is, a steady change state;

[0070] For KPSS test:

[0071] Set the null hypothesis H0 as: the extracted node data sequence is in a stable changing state;

[0072] The alternative hypothesis H1 is: the extracted node data sequence is in a non-stationary changing state;

[0073] Step S32: Perform an ADF test to obtain the test result of the ADF test. The specific process of the ADF test includes:

[0074] Calculate the ADF test statistic;

[0075] If the calculated p-value is less than the set significance level and the ADF test statistic is less than the set critical value, the null hypothesis is rejected and it is considered that the extracted node data sequence is in a stable change state;

[0076] If the calculated p-value is greater than the set significance level, or the ADF test statistic is greater than the set critical value, the null hypothesis is accepted and it is considered that the extracted node data sequence is in a non-stationary change state;

[0077] Step S33: Perform KPSS inspection and obtain the inspection result of KPSS inspection. The specific inspection process of KPSS inspection includes:

[0078] Calculate the KPSS test statistic;

[0079] If the calculated KPSS test statistic is less than the set critical value and the calculated p-value is greater than the significance level, the null hypothesis cannot be rejected and it is considered that the extracted node data sequence is in a stable change state;

[0080] If the calculated KPSS test statistic is greater than the set critical value, or the calculated p-value is less than the significance level, the original hypothesis of stationarity is rejected, and it is considered that the extracted node data sequence is in a non-stationary change state.

[0081] Step S34: Cross-validation results

[0082] When the results of the ADF test and the KPSS test are consistent

[0083] If both indicate that the extracted node data sequence is in a steady-changing state, then the extracted node data sequence is considered to be in a steady-changing state;

[0084] If both indicate that the extracted node data sequence is in a non-stationary change state, then the extracted node data sequence is considered to be in a non-stationary change state;

[0085] When the results of the ADF test and the KPSS test are inconsistent

[0086] Perform first-order or higher-order difference processing on the extracted node data series until the results of the ADF test and the KPSS test are consistent.

[0087] It should be noted that the significance level is a pre-set threshold in hypothesis testing that determines the criteria for rejecting the null hypothesis. It represents the probability threshold of observing the current data or more extreme data if the null hypothesis is true.

[0088] Common significance levels:

[0089] 0.05 (5%): The most commonly used significance level, indicating a 5% probability of making a Type I error (i.e., incorrectly rejecting the null hypothesis);

[0090] 0.01 (1%): A more stringent significance level, indicating a 1% probability of making a Type I error;

[0091] 0.10 (10%): A more relaxed significance level, indicating a 10% probability of making a Type I error.

[0092] The p-value is the probability of observing the current data or more extreme data if the null hypothesis is true. The smaller the p-value, the more inconsistent the data are with the null hypothesis, and the more reasonable it is to reject the null hypothesis. If the p-value is less than the significance level, the null hypothesis is rejected. If the p-value is greater than or equal to the significance level α, the null hypothesis is not rejected.

[0093] A test statistic is a value calculated based on sample data to measure the degree of difference between the sample data and the null hypothesis. Different hypothesis tests have different statistical calculation methods. Common statistics include:

[0094] Z statistic: used for large sample mean test, the calculation formula is: Where X is the sample mean; u0 is the hypothesized population mean; σ is the population standard deviation; and n is the sample size.

[0095] T-statistic: used for small sample mean test, the calculation formula is: Among them, among them, is the sample mean; u0 is the hypothesized population mean; σ is the population standard deviation; n is the sample size; and s is the sample standard deviation.

[0096] F statistic: used for analysis of variance (ANOVA) and regression analysis, the calculation formula is: F = MS between / MS within Among them, MS between is the between-group mean square; MS within is the within-group mean square.

[0097] Chi-square statistic: used for chi-square test, the calculation formula is: χ 2 =∑(OE) 2 / E; where O is the observed frequency and E is the expected frequency.

[0098] The critical value is a threshold value determined based on the significance level (α) and the distribution of the statistic. If the absolute value of the calculated statistic is greater than the critical value, the null hypothesis is rejected. The determination of the critical value requires consideration of the significance level, the distribution of the statistic, and the test type. Specifically, the critical value can usually be found using statistical tables or statistical software. The following is a specific example of how to determine the critical value based on search results:

[0099] Determine the significance level. In this scheme, the significance level is 0.05;

[0100] Select a statistical distribution

[0101] Choose an appropriate statistical distribution based on the test type, such as the standard normal distribution (Z distribution), t distribution, chi-square distribution, or F distribution;

[0102] Find or calculate critical values

[0103] Standard normal distribution (Z distribution):

[0104] For a two-tailed test, find the Z value that corresponds to half the significance level. For example, if the significance level is 0.05, find the Z value that corresponds to 0.025 (that is, 10.05 / 2).

[0105] t-distribution:

[0106] Use the t-distribution when the population standard deviation is unknown and the sample size is small. Find the t-critical value based on the significance level and the degrees of freedom (sample size minus 1).

[0107] Chi-square distribution:

[0108] For the chi-square test, find the critical value based on the significance level and degrees of freedom.

[0109] F-distribution:

[0110] Used in analysis of variance (ANOVA), etc., to find critical values ​​based on the significance level and the degrees of freedom of the numerator and denominator.

[0111] Using a tool or form

[0112] Critical value table: Traditional statistics textbooks or manuals usually come with a critical value table, which can be directly searched.

[0113] Online Calculator: Use the online critical value calculator to enter the significance level, degrees of freedom, and distribution type and get quick results.

[0114] Statistical software: Statistical software such as SPSS, R, Python, etc. can calculate critical values

[0115] Step 4: Based on the state classification results of the node data, a prediction model is constructed using the extracted node data. The prediction model includes a first prediction model when the node data is in a stable change state and a second prediction model when the node data is in a non-stationary change state;

[0116] Among them, when the node data is in a stable changing state, the first prediction model is constructed using the node data, and the expression is:

[0117]

[0118] Where, is the node prediction data of the operating state parameter obtained by the first prediction model in the period from t = q to t = q + 1; c is the constant coefficient; ε is the error coefficient; μk is the model coefficient; p is an integer between 1 and q-1;

[0119] When the node data is in a non-stationary change state, the second prediction model is constructed using the node data, and the expression is:

[0120]

[0121] Where, is the node prediction data of the operating state parameter obtained by the second prediction model in the period from t = q to t = q + 1; c is the constant coefficient; ε is the error coefficient; μ k is the model coefficient; p is an integer between 1 and q-1; ε (q-k+1) is the random error term; ν k is the random error coefficient.

[0122] Step 5: Substitute the extracted node data into the constructed prediction model, solve the model parameters, and obtain the node prediction data of the operating status parameters in the period from t=q to t=q+1.

[0123] Step 6: Add the obtained node prediction data to the first time series data sequence Rs i (t) forms the second time series data sequence Rs' i (t).

[0124] In this embodiment, the node data is in a stable changing state, and the node prediction data is obtained using the constructed first prediction model. Then the expression of the second time series data sequence can be written as: According to the above expression, the data volume of the data sequence reconstructed using this scheme is doubled compared to the data volume of the sampled data, so that the amount of data used to drive the digital twin model is doubled on the basis of unchanged data acquisition frequency, which is beneficial to improving the accuracy of real-time mapping between virtual models and physical entities.

[0125] It should also be noted that, in practical applications, the operating status parameter node data in the period from t=q to t=q+1 can be expanded according to the required data volume to meet the needs of model construction.

[0126] Step 7: Use the acquired second time series data sequence Rs' i (t) Drives the creation of a digital twin model for shipbuilding projects, through which the processing of the monitored processing equipment is virtually simulated and monitored in real time.

[0127] Among them, for the construction of digital twin models of shipbuilding projects, 3D modeling software (such as SolidWorks, 3dsMax, Unity3D) can be used to accurately model cutting equipment and welding equipment to ensure that the virtual model is highly consistent with the physical equipment in structure and function. The operating status parameters of the equipment are obtained as the second time series data sequence Rs' i (t)Integrated into the digital twin model.

[0128] The basic principles, main features, and advantages of the present invention are shown and described above. Those skilled in the art should understand that the present invention is not limited to the above embodiments. The above embodiments and descriptions are merely illustrative of the principles of the present invention. Various changes and modifications may be made to the present invention without departing from the spirit and scope of the present invention. Such changes and modifications are intended to fall within the scope of the present invention. The scope of protection claimed in the present invention is defined by the appended claims and their equivalents.

Claims

1. A real-time monitoring method for shipbuilding engineering, characterized in that: include: Step 1: Collect the operating status parameters of the monitored processing equipment during the processing at a sampling frequency of f, and construct the first time series data sequence Rs of the i-th operating status parameter i (t) = {rs i1 ,rs i2 ,...,rs in }, where t represents time; rs in is the sampling value of the operating status parameter when t=n; Step 2: Extract the first time series data sequence Rs i (t) The operating status parameter node data {rs i1 ,rs i2 ,...,rs iq }, using KPSS test and ADF test for cross validation, the change state of the extracted node data is determined, and the change state of the node data is classified as a stable change state or a non-stationary change state; Step 3: Based on the state classification result of the node data, a prediction model is constructed using the extracted node data, wherein the prediction model includes a first prediction model when the node data is in a stable change state and a second prediction model when the node data is in a non-stationary change state; Step 4: According to the constructed prediction model, obtain the node prediction data of the operating state parameter in the period from t=q to t=q+1, and add the obtained node prediction data to the first time series data sequence Rs i (t) forms the second time series data sequence Rs' i (t); Step 5: Use the acquired second time series data sequence Rs' i (t) Drives the creation of a digital twin model of the shipbuilding project, through which the processing of the monitored processing equipment is virtually simulated and monitored in real time.

2. A real-time monitoring method for shipbuilding engineering according to claim 1, characterized in that: The processing equipment includes cutting equipment and welding equipment, wherein: The operating state parameters of the cutting equipment include one or more combinations of cutting power, cutting speed, cutting air pressure and cutting head height; The operating status parameters of the welding equipment include one or more combinations of welding current, welding voltage, welding speed, wire consumption, gas flow, energy consumption value and welding quality parameters.

3. A real-time monitoring method for shipbuilding engineering according to claim 1, characterized in that: The specific process of step 2 includes the following steps: Step S21: Explicitly test the hypothesis For the ADF test: Set the null hypothesis H0 as: the extracted node data sequence has a unit root, that is, a non-stationary change state; The alternative hypothesis H1 is: there is no unit root in the extracted node data sequence, that is, a steady change state; For KPSS test: Set the null hypothesis H0 as: the extracted node data sequence is in a stable changing state; The alternative hypothesis H1 is: the extracted node data sequence is in a non-stationary changing state; Step S22: Perform ADF test and obtain the test result of ADF test Step S23: Perform KPSS test and obtain the test result of KPSS test Step S24: Cross-validation results When the results of the ADF test and the KPSS test are consistent If both indicate that the extracted node data sequence is in a steady-changing state, then the extracted node data sequence is considered to be in a steady-changing state; If both indicate that the extracted node data sequence is in a non-stationary change state, then the extracted node data sequence is considered to be in a non-stationary change state; When the results of the ADF test and the KPSS test are inconsistent Perform first-order or higher-order difference processing on the extracted node data series until the results of the ADF test and the KPSS test are consistent.

4. A real-time monitoring method for shipbuilding engineering according to claim 1, characterized in that: The expression of the first prediction model is: Where, is the node prediction data of the operating state parameter obtained by the first prediction model in the period from t = q to t = q + 1; c is the constant coefficient; ε is the error coefficient; μ k is the model coefficient; p is an integer between 1 and q-1.

5. A real-time monitoring method for shipbuilding engineering according to claim 1, characterized in that: The expression of the second prediction model is: Where, is the node prediction data of the operating state parameter obtained by the second prediction model in the period from t = q to t = q + 1; c is the constant coefficient; ε is the error coefficient; μ k is the model coefficient; p is an integer between 1 and q-1; ε (q-k+1) is the random error term; ν k is the random error coefficient.

6. A real-time monitoring method for shipbuilding engineering according to claim 3, characterized in that: The ADF inspection process includes: Calculate the ADF test statistic; If the calculated p-value is less than the set significance level and the ADF test statistic is less than the set critical value, the null hypothesis is rejected and it is considered that the extracted node data sequence is in a stable change state; If the calculated p-value is greater than the set significance level, or the ADF test statistic is greater than the set critical value, the null hypothesis is accepted and it is considered that the extracted node data sequence is in a non-stationary change state; The KPSS inspection process includes: Calculate the KPSS test statistic; If the calculated KPSS test statistic is less than the set critical value and the calculated p-value is greater than the significance level, the null hypothesis cannot be rejected and it is considered that the extracted node data sequence is in a stable change state; If the calculated KPSS test statistic is greater than the set critical value, or the calculated p-value is less than the significance level, the null hypothesis of stationarity is rejected, and it is considered that the extracted node data sequence is in a non-stationary change state.

7. A real-time monitoring method for shipbuilding engineering according to claim 1, characterized in that: The significance level was set at 0.

05.

8. A real-time monitoring system for shipbuilding projects, characterized in that: The system is used to implement the steps of the real-time monitoring method for shipbuilding projects according to any one of claims 1 to 7, including: The state parameter acquisition module is used to collect the operating state parameters of the monitored processing equipment during the processing at a sampling frequency of f, and construct the first time series data sequence Rs of the i-th operating state parameter i (t) = {rs i1 ,rs i2 ,...,rs in }, where t represents time; rs in is the sampling value of the operating status parameter when t=n; Data verification module, used to extract the first time series data sequence Rs i (t) The operating status parameter node data {rs i1 ,rs i2 ,...,rs iq }, using KPSS test and ADF test for cross validation, the change state of the extracted node data is determined, and the change state of the node data is classified as a stable change state or a non-stationary change state; A prediction model building module is used to build a prediction model using the extracted node data according to the state classification result of the node data, including a first prediction model when the node data is in a stable change state and a second prediction model when the node data is in a non-stationary change state; The data reconstruction module is used to obtain the node prediction data of the operating state parameter in the period from t=q to t=q+1 according to the constructed prediction model, and add the obtained node prediction data to the first time series data sequence Rs i (t) forms the second time series data sequence Rs' i (t); A digital twin model construction module is used to construct a digital twin model for virtual simulation and real-time monitoring of the processing process of the monitored processing equipment; Model driving module, used to obtain the second time series data sequence Rs' i (t) Importing it into the digital twin model, thereby driving the created digital twin model of the shipbuilding project to perform virtual simulation and real-time monitoring on the processing process of the monitored processing equipment.

9. A real-time monitoring system for shipbuilding projects according to claim 8, characterized in that: The system includes a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor can implement the steps of the method for real-time monitoring of shipbuilding projects according to any one of claims 1 to 7 when executing the program.

Images