Automatic monitoring system for plasma meltallizing coating process data

Abstract

The invention relates to the technical field of data monitoring, in particular to an automatic monitoring system for plasma meltallizing coating process data, which is characterized in that a parameter association relationship analysis module performs association analysis based on real-time parameters in a plasma meltallizing coating process, identifies an association relationship among the parameters and generates a process parameter association analysis result; according to the method, through multi-dimensional correlation analysis and path screening of the technological parameters, key factors influencing the coating quality and action links of the key factors are defined, and the abnormal fluctuation state and the transfer path can be accurately predicted in combination with parameter priority dynamic adjustment and real-time data feedback, so that the stability and continuity in the technological process are ensured. In the scheme, a dynamic weight distribution strategy is adopted, so that key parameters in an abnormal state occupy a higher priority position, parameter distribution is rebalanced when the normal state is recovered, and excessive interference of a single parameter on the whole process is avoided.

Description

Technical Field

[0001] The invention relates to the technical field of data monitoring, and in particular to an automatic monitoring system for plasma spray coating process data. Background Art

[0002] The field of data monitoring technology uses various sensors to obtain physical parameters or status data, communication technology to achieve data transmission, and computers or embedded devices to analyze and process data.

[0003] Among them, the plasma spray coating process data automation monitoring system refers to a system that monitors and manages a large amount of process data generated in the plasma spray coating process in real time. It covers the collection of key parameters such as temperature, pressure, flow rate, etc. in the spray coating process, uses precision sensor equipment as the data input end, extracts and analyzes the collected raw data through multi-dimensional data modeling, and realizes real-time monitoring of process stability through dynamic threshold judgment.

[0004] The existing technology has difficulty in fully identifying the deep relationship between parameters when dealing with process parameter associations, and lacks the ability to adjust the dynamic priority of parameters and path strength in real time. For example, when a certain parameter fluctuates greatly, the existing technology cannot adjust its priority and associated links in real time, which may cause interference with process stability. In terms of the transfer prediction of abnormal parameter states, the existing technology's analysis of abnormal fluctuation trends is only based on static data models, and it is difficult to capture the changing laws of abnormal states in time series. This deficiency may make it difficult to respond to potential threats to process quality in a timely manner. In addition, it is difficult for the existing technology to achieve precise alignment of parameters and process stages, and it is impossible to effectively identify the parameter change characteristics of each stage, which will make the direction of process optimization lack of pertinence. For example, in the fine adjustment of the spraying temperature, traditional methods may ignore the synergistic effect of the injection current and gas flow, resulting in uneven or frequent fluctuations in coating quality indicators. When integrating data feedback on coating quality and parameter execution effects, the existing technology also has deficiencies in the analysis of multidimensional data correlation, making it difficult to provide a complete optimization reference for the overall process. Summary of the invention

[0005] In order to solve the technical problems existing in the prior art, the embodiment of the present invention provides a plasma spray coating process data automatic monitoring system. The technical solution is as follows:

[0006] A plasma spray coating process data automatic monitoring system, the system comprising:

[0007] The parameter correlation analysis module performs correlation analysis based on the real-time parameters in the plasma spray coating process, identifies the correlation between the parameters, and generates process parameter correlation analysis results;

[0008] The association chain dynamic adjustment module performs real-time adjustment and priority analysis on key parameters involved in the association path based on the process parameter association analysis result, and generates a real-time adjusted association path;

[0009] The current process status real-time monitoring module monitors the changes of the key path in the real-time adjusted associated path, marks the abnormal fluctuations or deviations from the standard range of the key path, and generates a monitoring record of the current process status;

[0010] The process abnormal state prediction module analyzes the abnormal fluctuation state trend in the monitoring record of the current process state to perform prediction analysis and obtain the abnormal transfer state prediction result;

[0011] The process optimization reference module adjusts the weight ratio of each stage of the plasma spray coating process based on the abnormal transfer state prediction result, selects the optimal reference process parameters according to the adjustment result, and generates the process optimization parameter analysis result;

[0012] The process optimization execution module dynamically optimizes the real-time execution process of the plasma spray coating process based on the process optimization parameter analysis results, and integrates the information to obtain the whole process monitoring data.

[0013] The present invention has improvements in that the process parameter association analysis results include a parameter association matrix, a key parameter influence path, and a parameter time lag analysis; the real-time adjusted association path includes a dynamically adjusted key path, a parameter priority order, and a parameter influence range update result; the monitoring record of the current process state includes a key parameter state change record, an abnormal fluctuation mark, and a parameter record that deviates from the standard range; the abnormal transfer state prediction result includes an abnormal continuous state prediction, an abnormal regression state prediction, and a parameter transfer path prediction; the process optimization parameter analysis result includes an allocation weight parameter, an optimal reference process parameter, and a process stage alignment parameter; the full process monitoring data includes a parameter optimization execution effect monitoring record, a process optimization target achievement status, and adjusted parameter execution feedback data.

[0014] The present invention is improved in that the parameter association relationship analysis module comprises:

[0015] The parameter data acquisition submodule obtains multi-dimensional process data including gas flow, injection current, and spraying temperature based on the real-time parameters in the plasma spray coating process, performs time series processing on the collected real-time process parameters, and generates standardized process parameter data;

[0016] The parameter association analysis submodule calculates the association degree between the process parameters based on the standardized process parameter data, determines the direction of the relationship between the parameters, establishes a directed graph with the parameters as nodes and the association as weighted edges, and generates a process parameter association graph;

[0017] The parameter association path screening submodule analyzes the association paths between parameters based on the process parameter association diagram, screens the key paths that affect the coating quality, identifies the parameter influence links with the process objectives, and generates process parameter association analysis results.

[0018] The present invention is improved in that the degree of correlation between process parameters is calculated by using the formula:

[0019]

[0020] Get the mutual information value I between two process parameters X and Y XY , represents the degree of information correlation between two parameters;

[0021] Among them, ∑ x∈X is the sum of all discrete intervals of parameter X, ∑ y∈Y It is the sum of all discrete intervals of parameter Y. p(x,y) represents the joint probability value of parameters X and Y being in the target interval at the same time. p(x) represents the marginal probability value of parameter X, and p(y) represents the marginal probability value of parameter Y.

[0022] The present invention is improved in that the association chain dynamic adjustment module includes:

[0023] The critical path monitoring submodule monitors the parameter changes of the critical path in the process in real time based on the process parameter correlation analysis results, performs trend analysis on the monitored parameter data according to the real-time process data, identifies the fluctuation state of the parameters in the critical path, and generates critical path parameter state data;

[0024] The association range judgment submodule makes a real-time judgment on the influence range of the parameters in the critical path based on the critical path parameter status data, compares the parameter fluctuation status with the established influence range threshold, determines whether there is a change in the influence range according to the comparison result, re-evaluates the association strength of the critical path, and generates a critical path influence range evaluation result;

[0025] The path dynamic adjustment submodule dynamically adjusts the key path in the process based on the evaluation results of the critical path impact range, modifies the parameter priority and association strength of the path in real time, updates the association link according to the adjusted path, records and stores the real-time status of the current process association chain, and generates a real-time adjusted association path.

[0026] The present invention is improved in that the current process status real-time monitoring module comprises:

[0027] The parameter change detection submodule monitors the parameters in the critical path in real time based on the real-time adjusted associated path, collects the current state data of the critical parameters, obtains the trend and fluctuation range of the parameter changes through time series analysis, and generates the critical path parameter change data;

[0028] The parameter abnormality comparison submodule compares the current state value of the parameter in the critical path with the preset standard threshold range based on the critical path parameter change data, determines the abnormal state of the parameter by judging whether the deviation amplitude exceeds the threshold, classifies and marks the abnormal situation of each key parameter, and generates critical path parameter abnormality information;

[0029] Based on the critical path parameter abnormality information, the abnormal marking and recording submodule organizes all abnormal parameters and the corresponding deviation ranges, marks the occurrence time, parameter position and amplitude of the abnormal state, records the parameter states and changes of all abnormal fluctuations, and generates a monitoring record of the current process state.

[0030] The present invention is improved in that the process abnormal state prediction module comprises:

[0031] The abnormal trend analysis submodule performs time series trend analysis on the parameters of abnormal fluctuations based on the monitoring records of the current process state, obtains the rate and change direction of the abnormal fluctuations, extracts the change rules of the abnormal state at multiple times, and generates abnormal fluctuation data;

[0032] The future state prediction submodule predicts the change trajectory of future abnormal fluctuations based on the abnormal fluctuation data, combined with all process state records and parameter change rates, analyzes the transfer direction and potential impact path of the abnormal state, and generates abnormal fluctuation transfer direction data;

[0033] The state transition classification submodule determines whether the abnormal state persists or gradually returns to the normal range based on the abnormal fluctuation transfer direction data and the current state data of the parameters, classifies the transfer type of the abnormal state, and generates an abnormal transfer state prediction result.

[0034] The present invention has the following improvements: for predicting the changing trajectory of future abnormal fluctuations, the formula is adopted:

[0035]

[0036] Calculate from the current state Z t Transition to next state Z t+1 The transition probability Q(Z t+1 |Z t );

[0037] Among them, Z tIndicates the state category of the parameter at time point t, Z t+1 Indicates the state category of the parameter at time point t+1, M(Z t+1 ,Z t ) is the current state Z t and the next state Z t+1 The number of joint occurrences of t ) is the current state Z t The total number of occurrences of .

[0038] The present invention is improved in that the process optimization reference module comprises:

[0039] The parameter weight adjustment submodule adjusts the weight ratio of the plasma spray coating process parameters based on the abnormal transfer state prediction result. When the abnormal state is a continuous abnormality, the weight ratio of the abnormal parameters is increased. When the state returns to normal, the weight distribution is balanced between the recovery stage and the stable stage to generate the adjusted parameter weight.

[0040] The process stage alignment submodule aligns the optimized parameters with each stage in the current process based on the adjusted parameter weights, identifies the parameters and change characteristics of each process stage, and generates a process parameter identification result;

[0041] The optimal parameter screening submodule evaluates the parameter optimization combination based on the process parameter identification results, screens the parameter configuration that best affects the coating quality, and generates process optimization parameter analysis results.

[0042] The present invention is improved in that the process optimization execution module includes:

[0043] The parameter dynamic adjustment submodule dynamically adjusts the key process parameters of the plasma spray coating process based on the process optimization parameter analysis results, refines the gas flow rate, injection current, and spray temperature parameters according to the optimized weight ratio, and synchronously corrects the parameter change trend in combination with the real-time collected data to generate dynamic adjustment parameter data;

[0044] The execution effect monitoring submodule continuously monitors the execution effect of the adjusted parameters based on the dynamically adjusted parameter data, records the relationship between the process parameter changes and the coating quality indicators, and generates the adjusted effect data.

[0045] Based on the adjusted effect data, the target evaluation submodule compares and analyzes the monitoring results with the preset process optimization goals, evaluates the effect of the process adjustment, integrates the execution information of all adjusted parameters, and generates full-process monitoring data.

[0046] The beneficial effects brought about by the technical solution provided by the embodiment of the present invention include at least:

[0047] Through multi-dimensional correlation analysis and path screening of process parameters, the key factors affecting coating quality and their action links are identified. Combined with dynamic adjustment of parameter priority and real-time data feedback, the abnormal fluctuation state and transfer path can be accurately predicted, thus ensuring the stability and continuity of the process. The strategy of dynamic weight distribution is adopted, so that the key parameters under abnormal conditions occupy a higher priority, and the parameter distribution is rebalanced when the normal state is restored, avoiding excessive interference of a single parameter on the overall process. Through the evaluation and detailed adjustment of parameter optimization configuration, the accuracy of parameter configuration is effectively improved in each process stage, and the synergy between coating quality indicators and adjustment effects is improved simultaneously. Real-time data collection, processing and feedback run through the entire optimization process, so that the process state can be closed-loop optimized under data drive, avoiding the lag problem that may be caused by traditional static threshold setting, and fundamentally improving the efficiency and quality of the overall process. BRIEF DESCRIPTION OF THE DRAWINGS

[0048] In order to more clearly illustrate the technical solutions in the embodiments of the present invention, the following briefly introduces the drawings required for use in the description of the embodiments. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without creative work.

[0049] Figure 1 It is a system module diagram of the present invention;

[0050] Figure 2 It is a system framework diagram of the present invention;

[0051] Figure 3 Schematic diagram of the parameter association relationship analysis module of the present invention;

[0052] Figure 4 It is a schematic diagram of the association chain dynamic adjustment module of the present invention;

[0053] Figure 5 Schematic diagram of a real-time monitoring module for the current process status of the present invention;

[0054] Figure 6 Schematic diagram of a process abnormal state prediction module of the present invention;

[0055] Figure 7 A schematic diagram of a reference module for optimizing the process of the present invention;

[0056] Figure 8 It is a schematic diagram of the process optimization execution module of the present invention. DETAILED DESCRIPTION

[0057] The technical solution of the present invention is described below in conjunction with the accompanying drawings.

[0058] In the embodiments of the present invention, words such as "exemplarily" and "for example" are used to indicate examples, illustrations or explanations. Any embodiment or design described as "example" in the present invention should not be interpreted as being more preferred or more advantageous than other embodiments or designs. Specifically, the use of the word "example" is intended to present the concept in a specific way. In addition, in the embodiments of the present invention, the meaning expressed by "and / or" can be both, or it can be either of the two.

[0059] In the embodiments of the present invention, "image" and "picture" can sometimes be used interchangeably. It should be noted that when the difference between them is not emphasized, the meanings they intend to express are the same. "of", "corresponding, relevant" and "corresponding" can sometimes be used interchangeably. It should be noted that when the difference between them is not emphasized, the meanings they intend to express are the same.

[0060] In the embodiments of the present invention, sometimes a subscript such as W1 may be written as a non-subscript such as W1. When the difference is not emphasized, the meanings to be expressed are the same.

[0061] In order to make the technical problems, technical solutions and advantages to be solved by the present invention more clear, a detailed description will be given below with reference to the accompanying drawings and specific embodiments.

[0062] The embodiment of the present invention provides a plasma spray coating process data automatic monitoring system, the system comprising:

[0063] The parameter correlation analysis module performs correlation analysis based on the real-time parameters in the plasma spray coating process, identifies the correlation between the parameters, and generates process parameter correlation analysis results;

[0064] The dynamic adjustment module of the association chain performs real-time adjustment and priority analysis on the key parameters involved in the association path based on the results of the process parameter association analysis, and generates a real-time adjusted association path;

[0065] The current process status real-time monitoring module monitors the changes in the key path in the associated path after real-time adjustment, marks the abnormal fluctuations or deviations from the standard range of the key path, and generates a monitoring record of the current process status;

[0066] The process abnormal state prediction module analyzes the abnormal fluctuation state trend in the monitoring record of the current process state and performs prediction analysis to obtain the abnormal transfer state prediction result;

[0067] The process optimization reference module adjusts the weight ratio of each stage of the plasma spray coating process based on the abnormal transfer state prediction results, selects the optimal reference process parameters according to the adjustment results, and generates process optimization parameter analysis results;

[0068] The process optimization execution module dynamically optimizes the real-time execution process of the plasma spray coating process based on the process optimization parameter analysis results, and integrates the information to obtain the monitoring data of the entire process;

[0069] The results of process parameter association analysis include parameter association matrix, key parameter influence path, and parameter time lag analysis. The real-time adjusted association path includes the dynamically adjusted key path, parameter priority order, and parameter influence range update results. The monitoring records of the current process status include key parameter state change records, abnormal fluctuation marks, and parameter records that deviate from the standard range. The abnormal transfer state prediction results include abnormal continuous state prediction, abnormal regression state prediction, and parameter transfer path prediction. The process optimization parameter analysis results include allocation weight parameters, optimal reference process parameters, and process stage alignment parameters. The full process monitoring data includes parameter optimization execution effect monitoring records, process optimization target achievement status, and adjusted parameter execution feedback data.

[0070] See also Figure 2 and Figure 3 , the parameter association relationship analysis module includes:

[0071] The parameter data acquisition submodule obtains multi-dimensional process data including gas flow, injection current, and spraying temperature based on the real-time parameters in the plasma spray coating process, performs time series processing on the collected real-time process parameters, and generates standardized process parameter data;

[0072] Real-time signals are collected by industrial sensors. First, the noise of the collected original signals is filtered, and the high-frequency noise is decomposed and eliminated by wavelet transform, while the low-frequency signal is retained. Then, the filtered signals are arranged in chronological order to form the original time series data. For the missing values ​​that may exist in the data, the multivariate interpolation method is used to fill them. Specifically, the Lagrange interpolation method is used to estimate the change trend of adjacent points to fill the gaps. Further, the possible outliers in the data are detected, and the values ​​that deviate from the normal distribution range are eliminated by the 3 times standard deviation method. The eliminated data are re-interpolated and supplemented. Finally, the data is scaled to the interval between 0 and 1 by the normalization method to generate a standardized time series data matrix, and finally the standardized process parameter data is formed.

[0073] The parameter association analysis submodule calculates the degree of association between process parameters based on standardized process parameter data, determines the direction of the relationship between parameters, establishes a directed graph with parameters as nodes and association as weighted edges, and generates a process parameter association graph;

[0074] To calculate the correlation between process parameters, the formula is used:

[0075]

[0076] Get the mutual information value I between two process parameters X and Y XY , represents the degree of information correlation between two parameters;

[0077] Among them, ∑ x∈X It is to sum all the discrete intervals of parameter X to accumulate the contribution of the joint probability value of each interval in the entire parameter space. The symbol x∈X represents the value of all discrete intervals of parameter X. y∈Y It is to sum all the discrete intervals of parameter Y to accumulate the contribution of the joint probability value of each interval in the entire parameter space. The symbol y∈Y represents the value of parameter Y in all discrete intervals. p(x,y) represents the joint probability value of parameters X and Y in a certain interval at the same time. It is calculated by normalizing the joint frequency of the two parameter values. The formula is Where f(x i ,y j ) represents the frequency of X and Y in the corresponding interval, N is the total number of samples, i represents the interval index of parameter X, j represents the interval index of parameter Y, p(x) represents the marginal probability value of parameter X, which represents the probability that parameter X is alone in a certain interval, and is calculated by counting the frequency of the interval and normalizing it. The formula is Where m is the number of intervals of parameter Y, p(y) represents the marginal probability value of parameter Y, which means the probability that parameter Y is in a certain interval alone, and is calculated by counting the frequency normalization of the interval. The formula is: Where n is the number of intervals of parameter X, and log is a logarithmic function with natural logarithm as base, which is used to calculate the relative ratio of p(x,y) to p(x)p(y), indicating the relationship between the joint probability and marginal probability between parameters.

[0078] Parameter acquisition method:

[0079] Parameters X and Y: Two process parameters that need to be analyzed during the process, such as gas flow X and injection current Y. The values ​​of these parameters are collected through real-time monitoring by sensors. The parameters recorded by the sensors are stored in the form of time series, with one data point recorded per second.

[0080] Joint probability p(x,y): The frequency f(x) of two parameters appearing simultaneously in the corresponding interval in the statistically standardized time series data i ,y j), and then divided by the total number of samples N. For example, if the number of times the gas flow rate and the injection current are simultaneously in a certain interval is 20 among 100 sampling points, then the joint probability is

[0081]

[0082] Marginal probabilities p(x) and p(y): By counting the frequency f(x i ), and then divided by the total number of samples N. For example, if the total number of times the gas flow rate appears in a certain interval is 40, then the marginal probability is

[0083] If there are three process parameters: gas flow rate X1, spray current X2, and spray temperature X3. Each parameter is divided into 3 discrete intervals. An example of joint probability distribution is as follows: Joint probability distribution of gas flow rate X1 and spray current X2: When the gas flow rate is in interval 1, the joint probability of the spray current in interval 1 is 0.1, the joint probability in interval 2 is 0.05, and the joint probability in interval 3 is 0.1. When the gas flow rate is in interval 2, the joint probability of the spray current in interval 1 is 0.15, the joint probability in interval 2 is 0.1, and the joint probability in interval 3 is 0.05. When the gas flow rate is in interval 3, the joint probability of the spray current in interval 1 is 0.05, the joint probability in interval 2 is 0.05, and the joint probability in interval 3 is 0.1.

[0084] Marginal probability of gas flow X1:

[0085] p(X1 interval 1)=0.1+0.05+0.1=0.25.

[0086] p(X1 interval 2) = 0.15 + 0.1 + 0.05 = 0.3.

[0087] p(X1 interval 3)=0.05+0.05+0.1=0.2.

[0088] Edge probability of injection current X2:

[0089] p(X2 interval 1)=0.1+0.15+0.05=0.3.

[0090] p(X2 interval 2) = 0.05 + 0.1 + 0.05 = 0.2.

[0091] p(X2 interval 3)=0.1+0.05+0.1=0.25.

[0092] Substitute into the formula to calculate

[0093]

[0094] Calculate the mutual information value of interval 1 (x1, y1):

[0095] The joint probability p(x1,y1)=0.1.

[0096] Marginal probabilities p(x1)=0.25, p(y1)=0.3.

[0097]

[0098] Calculate the mutual information value of interval 2 (x1, y2):

[0099] Joint probability p(x1,y2)=0.05.

[0100] Marginal probabilities p(x1)=0.25, p(y2)=0.2.

[0101]

[0102] Calculate the mutual information value of interval 3(x2,y1):

[0103] The joint probability p(x2,y1)=0.15.

[0104] Marginal probabilities p(x2)=0.3, p(y1)=0.3.

[0105]

[0106] Accumulate the mutual information values ​​of all intervals:

[0107]

[0108] result This indicates that the mutual information value between gas flow rate and spray current is low and the correlation is weak. A similar method is used to calculate the mutual information values ​​between other process parameter pairs, such as gas flow rate and spraying temperature. Or spray current and spray temperature

[0109] Assume that the mutual information value range of process parameters is defined as follows: I XY >0.2: The correlation strength is high, and the parameters are closely related. 0.05≤I XY ≤0.2: The correlation strength is medium, and there may be an indirect relationship between the parameters. XY <0.05: The correlation strength is low and the relationship between the parameters is weak. According to the preset range, the current mutual information value Fall in I XY<0.05, so it can be judged that the correlation strength between gas flow rate and injection current is low. Further calculate the mutual information value of other process parameter pairs, such as gas flow rate and spraying temperature Spraying current and spraying temperature And also judge the strength of its association in combination with the preset range. For example, suppose the calculation results are and According to the scope definition: Falling within 0.05≤I XY In the range of ≤0.2, it can be judged that the correlation strength between gas flow rate and spraying temperature is medium, and there may be a certain indirect effect. Also falls within 0.05≤I XY The interval of ≤0.2 indicates that the correlation strength between the injection current and the spraying temperature is moderate. Next, the order of parameter changes is analyzed through time series data to determine the directionality of the relationship between the parameters. For example, through monitoring data, it is found that the change in gas flow rate usually precedes the change in spraying temperature, and the change in spraying temperature precedes the change in injection current, which indicates that the causal order between the parameters may be: gas flow rate → spraying temperature → injection current. On this basis, a directed graph is constructed with gas flow rate, injection current and spraying temperature as nodes, and the weight of the edge is the corresponding mutual information value. The specific construction process is as follows: Add nodes X1 (gas flow rate), X2 (injection current), and X3 (spraying temperature) to the graph. Add a directed edge from gas flow rate X1 to spraying temperature X3, with a weight of Add a directed edge from the spray temperature X3 to the spray current X2 with a weight of The directed graph constructed in this way clearly reflects the correlation strength and causal relationship between various parameters, providing a clear basis for subsequent process optimization and causal link analysis.

[0110] The parameter association path screening submodule analyzes the association paths between parameters based on the process parameter association diagram, screens the key paths that affect the coating quality, identifies the parameter influence links with the process objectives, and generates process parameter association analysis results;

[0111] First, all parameter nodes and directed edges between them are extracted. By counting the weight values ​​of directed edges and sorting them, the associated paths between the parameters are analyzed in turn. Paths with higher weight values ​​are screened first and identified as critical paths that may affect the coating quality. At the same time, associated edges with low weights are removed. According to the preset weight threshold range, for example, when the weight w>0.15, the path is marked as a critical path. If the weight is 0.05≤w≤0.15, it is marked as a secondary critical path. The path with weight w<0.05 is judged to have no significant effect and removed. In the path screening process, the parameter nodes in each path are further analyzed, such as calculating the cumulative weight of the path and the in-and-out degree of the node (the number of in-and-out edges of the node) to identify which parameter nodes have the most significant overall impact on the coating quality. The specific analysis method includes: for each node, the sum of its out-degree weight (the cumulative weight of all edges starting from the node) is calculated to measure the intensity of the influence of the parameter on other parameters. For each path, the cumulative weight of all edges on the path is calculated to evaluate the potential impact of the path on the coating quality. According to the process objectives, such as stable coating thickness or uniformity, identify which parameter nodes on the path are directly or indirectly related to the objectives. Suppose during the analysis, it is found that the path weight between gas flow and spray current is 0.18, the path weight between melting temperature and spray current is 0.1, and the path weight between gas flow and melting temperature is 0.25. According to the preset weight range, the path of gas flow → melting temperature is determined to be the critical path, while the path of melting temperature → spray current is the secondary critical path, and the path of gas flow → spray current is the secondary critical path. Finally, the screened critical paths are matched with the process objectives, and the key parameter nodes involved in the path are marked, such as marking gas flow and melting temperature as core parameters that have a significant impact on coating quality.

[0112] See also Figure 2 and Figure 4 ,The association chain dynamic adjustment module includes:

[0113] The critical path monitoring submodule monitors the parameter changes of the critical path in the process in real time based on the process parameter correlation analysis results. It performs trend analysis on the monitored parameter data according to the real-time process data, identifies the fluctuation status of the parameters in the critical path, and generates the critical path parameter status data.

[0114] Real-time monitoring of parameter changes in critical paths during the process. Real-time data of parameters such as gas flow, injection current, and spraying temperature are collected and stored through sensors. A dynamic change data set of each critical path parameter is constructed in the form of a time series. Differentiation operations are performed on the parameters in the dynamic change data set to obtain their short-term change rates, and their moving average values ​​are calculated for trend analysis. The parameter change trends are classified according to the preset range. For example, when the parameter change rate is lower than 0.05, it is judged as "stable", between 0.05 and 0.2 as "fluctuating", and higher than 0.2 as "violently fluctuating". After classification, the fluctuation status of all parameters in the critical path is marked, and finally the critical path parameter status data is output.

[0115] The association range judgment submodule makes a real-time judgment on the influence range of the parameters in the critical path based on the critical path parameter status data, compares the parameter fluctuation status with the established influence range threshold, determines whether there is a change in the influence range based on the comparison result, re-evaluates the association strength of the critical path, and generates a critical path influence range evaluation result;

[0116] By comparing the fluctuation state of each parameter in the current path with the set impact range threshold, the threshold interval of the impact range is defined. For example, when the critical path correlation strength is lower than 0.1, it is defined as "weak impact", between 0.1 and 0.3 as "medium impact", and higher than 0.3 as "strong impact". The correlation strength of each critical path is calculated in turn and it is determined whether its impact range has changed. For paths whose fluctuation state is higher than the established impact range threshold, their correlation strength is re-evaluated, and the overall strength of the path is calculated by accumulating the fluctuation weights of each parameter in the path. Combined with the above results, the critical path impact range assessment result is output.

[0117] The path dynamic adjustment submodule dynamically adjusts the key path in the process based on the critical path impact range assessment results, modifies the path parameter priority and association strength in real time, updates the association link according to the adjusted path, records and stores the real-time status of the current process association chain, and generates a real-time adjusted association path;

[0118] First, based on the evaluation data of the path before adjustment, key parameters with higher weights are selected to adjust their association strengths first. Then, the priorities of other related parameters are readjusted based on the real-time collected data. All adjusted paths are stored and marked. The association strengths of key paths are reallocated according to the adjustment priorities. The adjusted paths are mapped to new association links. At the same time, all association links in the real-time state are stored, and finally the real-time adjusted association paths are generated.

[0119] See also Figure 2 and Figure 5 , the current process status real-time monitoring module includes:

[0120] The parameter change detection submodule monitors the parameters in the critical path in real time based on the associated path adjusted in real time, collects the current status data of the key parameters, obtains the trend and fluctuation range of parameter changes through time series analysis, and generates the critical path parameter change data;

[0121] The current state data of key parameters, such as gas flow, injection current, and melting temperature, are collected through sensors deployed on process equipment. The collection frequency is set to once per second, and the data is stored in the form of time series. Trends are extracted from the collected data through time series analysis, and the moving average method is used to smooth the data to reduce the impact of short-term random fluctuations. For example, in the continuously sampled parameter data, the average value of the last five sampling points is taken as the current smoothed value to construct a trend curve of parameter changes. Differential analysis is performed on trend data to calculate the change ratio between consecutive time points. The change ratio is obtained by the ratio of the parameter value at the current time point to the parameter value at the previous time point, and the change ratio at each time point is used as the basis for quantifying the fluctuation range. When classifying the fluctuation range, the fluctuation range is divided into three levels: stable, slight fluctuation, and violent fluctuation. For example, when the fluctuation range is less than 5%, it is classified as stable, in the range of 5% to 15%, it is classified as slight fluctuation, and when it exceeds 15%, it is classified as violent fluctuation. Finally, according to the classification results, the fluctuation state of each parameter is marked to generate key path parameter change data.

[0122] The parameter anomaly comparison submodule compares the current state value of the parameter in the critical path with the preset standard threshold range based on the critical path parameter change data, determines the abnormal state of the parameter by judging whether the deviation amplitude exceeds the threshold, classifies and marks the abnormal situation of each key parameter, and generates critical path parameter anomaly information;

[0123] First, the standard range is set according to the parameter type. For example, the standard range of gas flow is set between 30 and 70, the target value is 50, and the standard range of spraying temperature is set between 2000 and 3000, and the target value is 2500. Compare the current value of each parameter in the critical path with its corresponding target value, and calculate the deviation from the target value. For example, when the current value of the gas flow is 40, its deviation is obtained by the ratio of the absolute value of the difference between the current value and the target value to the target value, and the result is 20%. According to the result of the deviation, the abnormal state is divided into three levels: normal range, mild abnormality and severe abnormality. The deviation is less than 10% marked as normal range, the deviation is between 10% and 20% marked as mild abnormality, and the deviation exceeds 20% marked as severe abnormality. Through data processing of all critical path parameters, the abnormal state is classified and sorted by parameter category, and the abnormal type and level of each parameter are marked, and finally the critical path parameter abnormality information is generated.

[0124] The abnormal marking and recording submodule sorts out all abnormal parameters and corresponding deviation ranges based on the abnormal information of key path parameters, marks the occurrence time, parameter location and amplitude of the abnormal state, records the parameter states and changes of all abnormal fluctuations, and generates monitoring records of the current process state;

[0125] Match the time series data with the fluctuation data of abnormal parameters, mark the time point when the abnormality occurs, and arrange the abnormal parameters in chronological order, such as marking a 15% deviation of the gas flow at a certain moment and a 25% deviation of the injection current at another moment, and record the time and parameter fluctuation amplitude in a table respectively. Classify the abnormal fluctuations according to the fluctuation amplitude, and extract the duration and fluctuation frequency of multiple fluctuations of the same parameter, and count the number of abnormal occurrences of each parameter in a given time period. Arrange the classified abnormal data according to time, parameter and deviation amplitude to form a monitoring record containing the complete characteristics of each abnormal parameter, such as recording that the fluctuation amplitude of a parameter in a certain time period is slightly abnormal, lasts for 10 seconds, and fluctuates twice, and finally generates a monitoring record of the current process status.

[0126] See also Figure 2 and Figure 6 , the process abnormal state prediction module includes:

[0127] The abnormal trend analysis submodule performs time series trend analysis on the parameters of abnormal fluctuations based on the monitoring records of the current process status, obtains the rate and change direction of abnormal fluctuations, extracts the changing rules of abnormal states at multiple times, and generates abnormal fluctuation data;

[0128] By extracting the values ​​of abnormal parameters at different time points, we construct time series data, and use the sliding window method to smooth the time series to obtain the rate and change direction of abnormal fluctuations. The rate is obtained by the difference of parameter values ​​at consecutive time points, that is, the difference between the parameter value at the current time point and the parameter value at the previous time point divided by the time interval; the change direction is marked by comparing the positive and negative changes of the parameters at two consecutive points, for example, an increase in parameter value is marked as "positive", a decrease in parameter value is marked as "negative", and a change amplitude close to zero is marked as "no significant change". The change rules of abnormal states at multiple times are extracted, and by analyzing the distribution of fluctuation rates, it is determined whether the fluctuation changes tend to weaken, strengthen, or remain stable. Finally, the time series analysis results are integrated into abnormal fluctuation data.

[0129] The future state prediction submodule is based on abnormal fluctuation data, combined with all process state records and parameter change rates, to predict the change trajectory of future abnormal fluctuations, analyze the transfer direction and potential impact path of abnormal states, and generate abnormal fluctuation transfer direction data;

[0130] To predict the changing trajectory of future abnormal fluctuations, the formula is used:

[0131]

[0132] Calculate from the current state Z t Transition to next state Z t+1 The transition probability Q(Z t+1 |Z t );

[0133] Among them, Z t is the abnormal state at the current time point, indicating the state category of the parameter at time point t, such as "normal", "mild abnormal", "serious abnormal", etc. By monitoring the parameter fluctuation values ​​recorded, the state is marked according to the preset abnormal classification range. For example, a parameter deviation of less than 10% is "normal", a deviation of 10% to 20% is "mild abnormal", and a deviation of more than 20% is "serious abnormal". t+1 is the abnormal state at the next time point, indicating the state category of the parameter at time point t+1, such as from "mild abnormality" to "normal". Similarly, the corresponding state is marked according to the classification standard through the parameter fluctuation value in the time series. M(Z t+1 ,Z t ) is the current state Z t and the next state Z t+1 The joint occurrence frequency of two consecutive time points in the time series is obtained by counting the combination of each pair of consecutive states in the time series, for example, the number of times time point t is "slightly abnormal" and time point t+1 is "normal". t ) is the current state Z t The total number of occurrences of state Z in the time series t The cumulative frequency of occurrence at all time points counts the total number of times each state occurs in the entire time series, for example, the number of times the "mild abnormality" state appears in the time series is counted.

[0134] In a process status monitoring data, the abnormal status at the following time points are recorded: "normal", "mild abnormal", "mild abnormal", "normal", "serious abnormal", "mild abnormal", "normal".

[0135] According to the defined classification range: deviation less than 10% is "normal". Deviation between 10% and 20% is "mild abnormal". Deviation greater than 20% is "serious abnormal". According to the monitoring records, the status of each time point is marked to obtain the sequence.

[0136] Statistical transfer times: The number of times "mild abnormality" was transferred to "normal" was 2 times. The number of times "mild abnormality" was transferred to "mild abnormality" was 1 time. The total number of times the "mild abnormality" state appeared was 3 times.

[0137] Calculate the transition probability:

[0138] The probability of "mild abnormality" transferring to "normal" is:

[0139]

[0140] The probability of "mild abnormality" transferring to "mild abnormality" is:

[0141]

[0142] The calculation results show that the probability of transferring from the "mild abnormality" state to "normal" is 66.6%, while the probability of maintaining the "mild abnormality" state is 33.3%.

[0143] The state transition classification submodule determines whether the abnormal state persists or gradually returns to the normal range based on the abnormal fluctuation transfer direction data and the current state data of the parameters, classifies the transfer type of the abnormal state, and generates the abnormal transfer state prediction result;

[0144] First, the probability distribution of the abnormal fluctuation transfer direction is analyzed, and the transfer probability of each state is summarized according to the transfer direction. For example, the probability of transferring from "mild abnormality" to "normal" is 66.6%, and the probability of maintaining "mild abnormality" as "mild abnormality" is 33.3%. The transfer direction of each state is marked and classified. By analyzing the distribution of the transfer probability, it is judged whether the abnormal state persists or gradually returns to the normal range. For example, when the self-circulation probability of the abnormal state is greater than 80%, it is marked as "persistent abnormality", and when the cumulative probability of transferring to the normal state is greater than 50%, it is marked as "regression trend". Combined with the current state data of the parameters, the timing of state transition is further refined, and the transition type is classified by duration and frequency. For example, for the state of continuous abnormality, the continuous time period in which the state remains unchanged is determined as short-term or long-term abnormality, and the threshold of short-term abnormality is set to less than 3 consecutive time points, and the threshold of long-term abnormality is more than 3 consecutive time points; for the state of gradual regression, by observing the frequency trend of transfer to normal, it is classified as fast regression (the frequency of transfer to normal is greater than 70%) or slow regression (the frequency is less than 70% but greater than 50%). According to the classification standard, the transfer type of all abnormal states is marked, and the data is sorted by state, transfer direction, and time period. The results are summarized in the form of an abnormal state classification table to generate the abnormal transfer state prediction results.

[0145] See also Figure 2 and Figure 7, the process optimization reference module includes:

[0146] The parameter weight adjustment submodule adjusts the weight ratio of the plasma spray coating process parameters based on the abnormal transfer state prediction results. When the abnormal state is a continuous abnormality, the weight ratio of the abnormal parameters is increased. When the state returns to normal, the weight distribution is balanced between the recovery stage and the stable stage to generate the adjusted parameter weights.

[0147] First, analyze the abnormal transfer state category of each parameter, for example, mark the abnormal state as "persistent abnormality" or "return to normal", and the corresponding parameters are gas flow, injection current, spraying temperature, etc. Set different weight adjustment rules according to the state category of the parameter. For example, when the abnormal state is a continuous abnormality, the weight ratio of the corresponding parameter increases by 20%. When the state returns to normal, the weight balance is distributed between the recovery stage (60%) and the stable stage (40%). Weight adjustment is achieved through the following steps: extract the initial weight value of the parameter, adjust the weight ratio in combination with the state category, and normalize the weights of all parameters so that the sum of the weights is 1. Weight normalization is obtained by dividing the adjusted weight of each parameter by the sum of the adjusted weights of all parameters. Finally, the adjusted weight of each parameter is recorded to generate the adjusted parameter weight.

[0148] The process stage alignment submodule aligns the optimized parameters with each stage in the current process based on the adjusted parameter weights, identifies the parameters and change characteristics of each process stage, and generates process parameter identification results;

[0149] Extract the parameter characteristics of each process stage. For example, the plasma spray coating process can be divided into three stages: preheating, coating deposition, and cooling. Assign the adjusted parameter weights to each process stage, calculate the parameter priority of each stage through the stage weight matching formula, compare the parameter optimization results with the stage process goals, and identify the key parameters of the stage and their changing characteristics. Classify the parameter characteristics of each stage. For example, in the coating deposition stage, the parameters with higher weights are the injection current and spraying temperature, and their changing characteristics are that the fluctuation range is concentrated within 10% above and below the target value; in the cooling stage, the key parameter is the gas flow rate, and its fluctuation range is small. By sorting out the parameter characteristics of each stage, a process parameter identification result containing the matching relationship between the stage and parameter characteristics is generated.

[0150] The optimal parameter screening submodule evaluates the parameter optimization combination based on the process parameter identification results, screens the parameter configuration that best affects the coating quality, and generates process optimization parameter analysis results;

[0151] Extract the characteristic values ​​of each parameter configuration, such as the target value and actual deviation value of the injection current, and calculate the impact of the deviation value on the coating uniformity and adhesion; analyze the historical performance data of each parameter configuration, such as by extracting the standard deviation value of the coating thickness uniformity and the failure stress value of the adhesion, to evaluate the actual impact of the parameter configuration. Compare the performance values ​​of different parameter configurations. For example, when the target value of the injection current is 600A, the standard deviation value of the uniformity is less than 5%, and the failure stress value reaches 50MPa, while the uniformity and adhesion are both deteriorated when the deviation value is 50A, then select the configuration with a target value of 600A. Integrate the preferred values ​​of each parameter into an overall configuration, generate an optimized combination of the optimal parameters, and form the process optimization parameter analysis results.

[0152] See also Figure 2 and Figure 8 , the process optimization execution module includes:

[0153] The parameter dynamic adjustment submodule dynamically adjusts the key process parameters of the plasma spray coating process based on the process optimization parameter analysis results, and fine-tunes the gas flow, injection current, and spray temperature parameters according to the optimized weight ratio. It also synchronously corrects the parameter change trend based on the real-time collected data to generate dynamic adjustment parameter data.

[0154] First, extract the target value range of each parameter in the optimization result, such as gas flow rate between 12-15L / min, injection current between 500-550A, and melting temperature between 3000-3200K. Combined with the real-time collected process data, for example, the current gas flow rate is 13L / min, the injection current is 520A, and the melting temperature is 3100K, calculate the change trend of each parameter through time series analysis, compare the current value with the target range, and judge the parameter deviation. For example, the gas flow deviation is 1%, the injection current deviation is 4%, and the melting temperature deviation is 3%. Dynamic adjustment is made according to the degree of deviation, for example, the injection current is adjusted to the median of the target range (525A), and the melting temperature is adjusted to the upper limit of the target range (3200K). New parameter values ​​are generated by dynamic updating of the adjusted parameters and recorded as dynamically adjusted parameter data.

[0155] The execution effect monitoring submodule continuously monitors the execution effect of the adjusted parameters based on the dynamic adjustment parameter data, records the relationship between the process parameter changes and the coating quality indicators, and generates the adjusted effect data;

[0156] The monitoring data includes real-time collection of coating thickness, adhesion and uniformity, such as the collected thickness value of 200-210μm, adhesion of 48-52MPa, and uniformity standard deviation of 5%. By correlating the parameters before and after adjustment with the quality indicators, for example, after the injection current is adjusted from 520A to 525A, the thickness standard deviation drops from 8% to 5%, and the adhesion increases to 50MPa, indicating that the adjustment of the injection current improves the uniformity and adhesion of the coating. During continuous monitoring, a monitoring cycle with a time interval of 5 minutes is used to record the parameters and corresponding quality indicators after each adjustment, and all related data are organized into adjusted effect data.

[0157] The target evaluation submodule compares and analyzes the monitoring results with the preset process optimization goals based on the adjusted effect data, evaluates the effect of the process adjustment, integrates the execution information of all adjusted parameters, and generates the monitoring data of the whole process;

[0158] Compare and analyze the monitoring results with the preset process optimization goals to evaluate the effectiveness of the process adjustment. The goals include coating thickness deviation less than 5%, adhesion greater than 50MPa, and uniformity standard deviation less than 5%. By comparing the adjusted data, for example, the adjusted thickness deviation is 4%, the adhesion reaches 50MPa, and the uniformity standard deviation is 5%, it is judged that the goal has been achieved. For the unachieved part, such as the thickness deviation is still between 4% and 5%, by back-tracing the adjusted parameter data, analyze the reasons for not achieving it, such as the fluctuation range of the injection current is still large, and it is recommended to further reduce the fluctuation range of the current to ±3A. Integrate all adjusted parameter execution information, including the final value of each parameter and the corresponding quality indicators, to generate full-process monitoring data including the target achievement status.

[0159] It should be understood that the term "and / or" in this article is only a description of the association relationship of associated objects, indicating that there can be three relationships. For example, A and / or B can represent: A exists alone, A and B exist at the same time, and B exists alone. A and B can be singular or plural. In addition, the character " / " in this article generally indicates that the associated objects before and after are in an "or" relationship, but it may also indicate an "and / or" relationship. Please refer to the context for specific understanding.

[0160] In the present invention, "at least one" means one or more, and "more than one" means two or more. "At least one of the following" or similar expressions refers to any combination of these items, including any combination of single or plural items. For example, at least one of a, b, or c can be represented by: a, b, c, ab, ac, bc, or abc, where a, b, c can be single or multiple.

[0161] It should be understood that in various embodiments of the present invention, the size of the serial numbers of the above-mentioned processes does not mean the order of execution. The execution order of each process should be determined by its function and internal logic, and should not constitute any limitation on the implementation process of the embodiments of the present invention.

[0162] Those skilled in the art will appreciate that the units and algorithm steps of each example described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are performed in hardware or software depends on the specific application and design constraints of the technical solution. Professional and technical personnel can use different methods to implement the described functions for each specific application, but such implementation should not be considered to be beyond the scope of the present invention.

[0163] Those skilled in the art can clearly understand that, for the convenience and brevity of description, the specific working processes of the above-described equipment, devices and units can refer to the corresponding processes in the aforementioned method embodiments and will not be repeated here.

[0164] In the several embodiments provided by the present invention, it should be understood that the disclosed devices, apparatuses and methods can be implemented in other ways. For example, the device embodiments described above are only schematic. For example, the division of the units is only a logical function division. There may be other division methods in actual implementation, such as multiple units or components can be combined or integrated into another device, or some features can be ignored or not executed. Another point is that the mutual coupling or direct coupling or communication connection shown or discussed can be through some interfaces, indirect coupling or communication connection of devices or units, which can be electrical, mechanical or other forms.

[0165] The units described as separate components may or may not be physically separated, and the components shown as units may or may not be physical units, that is, they may be located in one place or distributed on multiple network units. Some or all of the units may be selected according to actual needs to achieve the purpose of the solution of this embodiment.

[0166] In addition, each functional unit in each embodiment of the present invention may be integrated into one processing unit, or each unit may exist physically separately, or two or more units may be integrated into one unit.

[0167] If the functions are implemented in the form of software functional units and sold or used as independent products, they can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present invention can be essentially or partly embodied in the form of a software product that contributes to the prior art. The computer software product is stored in a storage medium and includes several instructions for enabling a computer device (which can be a personal computer, a server, or a network device, etc.) to perform all or part of the steps of the methods described in various embodiments of the present invention. The aforementioned storage medium includes: various media that can store program codes, such as a USB flash drive, a mobile hard disk, a read-only memory (ROM), a random access memory (RAM), a magnetic disk or an optical disk.

[0168] The above is only a specific embodiment of the present invention, but the protection scope of the present invention is not limited thereto. Any person skilled in the art can easily think of changes or substitutions within the technical scope disclosed by the present invention, which should be included in the protection scope of the present invention. Therefore, the protection scope of the present invention should be based on the protection scope of the claims.

Claims

1. The plasma spray coating process data automatic monitoring system is characterized by: The system comprises: The parameter correlation analysis module performs correlation analysis based on the real-time parameters in the plasma spray coating process, identifies the correlation between the parameters, and generates process parameter correlation analysis results; The association chain dynamic adjustment module performs real-time adjustment and priority analysis on key parameters involved in the association path based on the process parameter association analysis result, and generates a real-time adjusted association path; The current process status real-time monitoring module monitors the changes in the key path in the real-time adjusted associated path, marks the abnormal fluctuations or deviations from the standard range of the key path, and generates a monitoring record of the current process status; The process abnormal state prediction module analyzes the abnormal fluctuation state trend in the monitoring record of the current process state to perform prediction analysis and obtain the abnormal transfer state prediction result; The process optimization reference module adjusts the weight ratio of each stage of the plasma spray coating process based on the abnormal transfer state prediction result, selects the optimal reference process parameters according to the adjustment result, and generates the process optimization parameter analysis result; The process optimization execution module dynamically optimizes the real-time execution process of the plasma spray coating process based on the process optimization parameter analysis results, and integrates the information to obtain the whole process monitoring data.

2. The plasma spray coating process data automatic monitoring system according to claim 1 is characterized in that: The process parameter association analysis results include parameter association matrix, key parameter influence path, and parameter time lag analysis; the real-time adjusted association path includes dynamically adjusted key path, parameter priority order, and parameter influence range update results; the monitoring record of the current process state includes key parameter state change record, abnormal fluctuation mark, and parameter record that deviates from the standard range; the abnormal transfer state prediction result includes abnormal continuous state prediction, abnormal regression state prediction, and parameter transfer path prediction; the process optimization parameter analysis results include allocation weight parameters, optimal reference process parameters, and process stage alignment parameters; the full process monitoring data includes parameter optimization execution effect monitoring records, process optimization target achievement status, and adjusted parameter execution feedback data.

3. The plasma spray coating process data automatic monitoring system according to claim 1 is characterized in that: The parameter association relationship analysis module includes: The parameter data acquisition submodule obtains multi-dimensional process data including gas flow, injection current, and spraying temperature based on the real-time parameters in the plasma spray coating process, performs time series processing on the collected real-time process parameters, and generates standardized process parameter data; The parameter association analysis submodule calculates the association degree between the process parameters based on the standardized process parameter data, determines the direction of the relationship between the parameters, establishes a directed graph with the parameters as nodes and the association as weighted edges, and generates a process parameter association graph; The parameter association path screening submodule analyzes the association paths between parameters based on the process parameter association diagram, screens the key paths that affect the coating quality, identifies the parameter influence links with the process objectives, and generates process parameter association analysis results.

4. The plasma spray coating process data automatic monitoring system according to claim 3 is characterized in that: To calculate the correlation between process parameters, the formula is used: Get the mutual information value I between two process parameters X and Y XY , represents the degree of information correlation between two parameters; Among them, ∑ x∈X is the sum of all discrete intervals of parameter X, ∑ y∈Y is the sum of all discrete intervals of parameter Y, p ( x,y ) Indicates the joint probability value of parameters X and Y being in the target interval at the same time, p ( x ) Represents the marginal probability value of parameter X, p ( y ) Represents the marginal probability value of parameter Y.

5. The plasma spray coating process data automatic monitoring system according to claim 1 is characterized by: The association chain dynamic adjustment module includes: The critical path monitoring submodule monitors the parameter changes of the critical path in the process in real time based on the process parameter correlation analysis results, performs trend analysis on the monitored parameter data according to the real-time process data, identifies the fluctuation state of the parameters in the critical path, and generates critical path parameter state data; The association range judgment submodule makes a real-time judgment on the influence range of the parameters in the critical path based on the critical path parameter status data, compares the parameter fluctuation status with the established influence range threshold, determines whether there is a change in the influence range according to the comparison result, re-evaluates the association strength of the critical path, and generates a critical path influence range evaluation result; The path dynamic adjustment submodule dynamically adjusts the key path in the process based on the evaluation results of the critical path impact range, modifies the parameter priority and association strength of the path in real time, updates the association link according to the adjusted path, records and stores the real-time status of the current process association chain, and generates a real-time adjusted association path.

6. The plasma spray coating process data automatic monitoring system according to claim 1 is characterized by: The current process status real-time monitoring module includes: The parameter change detection submodule monitors the parameters in the critical path in real time based on the real-time adjusted associated path, collects the current state data of the critical parameters, obtains the trend and fluctuation range of the parameter changes through time series analysis, and generates the critical path parameter change data; The parameter abnormality comparison submodule compares the current state value of the parameter in the critical path with the preset standard threshold range based on the critical path parameter change data, determines the abnormal state of the parameter by judging whether the deviation amplitude exceeds the threshold, classifies and marks the abnormal situation of each key parameter, and generates critical path parameter abnormality information; Based on the critical path parameter abnormality information, the abnormal marking and recording submodule organizes all abnormal parameters and the corresponding deviation ranges, marks the occurrence time, parameter position and amplitude of the abnormal state, records the parameter states and changes of all abnormal fluctuations, and generates a monitoring record of the current process state.

7. The plasma spray coating process data automatic monitoring system according to claim 1 is characterized by: The process abnormal state prediction module includes: The abnormal trend analysis submodule performs time series trend analysis on the parameters of abnormal fluctuations based on the monitoring records of the current process state, obtains the rate and change direction of the abnormal fluctuations, extracts the change rules of the abnormal state at multiple times, and generates abnormal fluctuation data; The future state prediction submodule predicts the change trajectory of future abnormal fluctuations based on the abnormal fluctuation data, combined with all process state records and parameter change rates, analyzes the transfer direction and potential impact path of the abnormal state, and generates abnormal fluctuation transfer direction data; The state transition classification submodule determines whether the abnormal state persists or gradually returns to the normal range based on the abnormal fluctuation transfer direction data and the current state data of the parameters, classifies the transfer type of the abnormal state, and generates an abnormal transfer state prediction result.

8. The plasma spray coating process data automatic monitoring system according to claim 7 is characterized in that: To predict the changing trajectory of future abnormal fluctuations, the formula is used: Calculate from the current state Z t Transition to next state Z t+1 The transition probability Q ( Z t+1 |Z t) ; Among them, Z t Indicates the state category of the parameter at time point t, Z t+1 Indicates the state category of the parameter at time point t+1, M ( Z t+1 ,Z t) is the current state Z t and the next state Z t+1 The number of joint occurrences of ( Z t is the current state Z t The total number of occurrences of .

9. The plasma spray coating process data automatic monitoring system according to claim 1 is characterized by: The process optimization reference module includes: The parameter weight adjustment submodule adjusts the weight ratio of the plasma spray coating process parameters based on the abnormal transfer state prediction result. When the abnormal state is a continuous abnormality, the weight ratio of the abnormal parameters is increased. When the state returns to normal, the weight distribution is balanced between the recovery stage and the stable stage to generate the adjusted parameter weight. The process stage alignment submodule aligns the optimized parameters with each stage in the current process based on the adjusted parameter weights, identifies the parameters and change characteristics of each process stage, and generates a process parameter identification result; The optimal parameter screening submodule evaluates the parameter optimization combination based on the process parameter identification results, screens the parameter configuration that best affects the coating quality, and generates process optimization parameter analysis results.

10. The plasma spray coating process data automatic monitoring system according to claim 1, characterized in that: The process optimization execution module includes: The parameter dynamic adjustment submodule dynamically adjusts the key process parameters of the plasma spray coating process based on the process optimization parameter analysis results, refines the gas flow rate, injection current, and spray temperature parameters according to the optimized weight ratio, and synchronously corrects the parameter change trend in combination with the real-time collected data to generate dynamic adjustment parameter data; The execution effect monitoring submodule continuously monitors the execution effect of the adjusted parameters based on the dynamically adjusted parameter data, records the relationship between the process parameter changes and the coating quality indicators, and generates the adjusted effect data. Based on the adjusted effect data, the target evaluation submodule compares and analyzes the monitoring results with the preset process optimization goals, evaluates the effect of the process adjustment, integrates the execution information of all adjusted parameters, and generates full-process monitoring data.

Images