Method and system for detecting qualified rate of finished product of caprylic / capric triglyceride

Abstract

The invention relates to a finished product qualification rate detection method and system for caprylic / capric triglyceride, and relates to the technical field of caprylic / capric triglyceride detection, and the system comprises a sampling device, a conveying pipeline device and a shunting device; the flow dividing device is used for dividing the caprylic / capric triglyceride sample into a first branch flow, a second branch flow and a third branch flow; the caprylic / capric triglyceride sample of the first branch flows into a density detection module; the caprylic / capric triglyceride sample of the second branch flows into the dielectric constant detection module; the caprylic / capric triglyceride sample of the third branch flows into the chromatographic analysis module; the control module is electrically connected with the density detection module, the dielectric constant detection module and the chromatographic analysis module; and the control module obtains the comprehensive qualification index of the sample based on the sample dielectric constant, the sample density, the glyceride substance analysis result and the trained quality analysis model.

Description

Technical Field

[0001] The present application relates to the technical field of caprylic / decanoic acid triglyceride detection, and in particular to a method and system for detecting the qualified rate of finished products of caprylic / decanoic acid triglyceride. Background Art

[0002] As an important chemical product, caprylic / capric triglyceride is widely used in many fields such as food, medicine, and cosmetics.

[0003] In the prior art, the quality inspection of the finished product of caprylic acid triglyceride often faces many challenges and limitations. Traditional chemical product quality inspection mostly relies on manual regular sampling, which is then sent to a special laboratory for offline analysis and inspection. For caprylic acid triglyceride, this method has obvious lag, and the time interval from sampling to obtaining the test results is long, making it difficult to timely feedback the real-time quality of the products on the production line. Once a quality problem occurs, a large number of unqualified products may have been produced.

[0004] How to conduct the finished product qualification rate detection of caprylic / decanoic acid triglyceride more timely is a technical problem that technicians in this field need to overcome. Summary of the invention

[0005] In order to at least partially solve the above technical problems, the present application provides a method and system for detecting the qualified rate of finished products of caprylic / decanoic acid triglyceride.

[0006] In a first aspect, the present application provides a finished product qualification rate detection system for caprylic / decanoic acid triglyceride adopts the following technical solution.

[0007] A finished product qualification rate detection system for caprylic / decanoic acid triglyceride, characterized by comprising: A sampling device, used for sampling the finished product of caprylic / decanoic acid triglyceride on the production line; A delivery pipeline device connected to the sampling device; the delivery pipeline device is used to deliver the sampled caprylic / capric triglyceride sample; A flow dividing device is connected to the delivery pipeline device; the flow dividing device divides the caprylic / capric triglyceride sample into a first branch flow, a second branch flow and a third branch flow; Density detection module, the caprylic / capric triglyceride sample of the first branch flows into the density detection module; the density detection module is used to measure the density of the sample; A dielectric constant detection module, the caprylic / capric triglyceride sample of the second branch flows into the dielectric constant detection module; the dielectric constant detection module is used to measure the dielectric constant of the sample; Chromatographic analysis module, the caprylic / capric triglyceride sample of the third branch flows into the chromatographic analysis module; the chromatographic analysis module is used to analyze the glyceride substances in the sample; A control module, wherein the control module is electrically connected to the density detection module, the dielectric constant detection module and the chromatographic analysis module; the control module obtains a comprehensive qualification index of the sample based on the sample dielectric constant, the sample density, the analysis results of the glyceride substances and the trained quality analysis model; when the comprehensive qualification index is lower than a preset threshold, the caprylic / capric triglyceride sample is determined to be unqualified, and the control module obtains the suspected cause of the failure in the production link according to the detection data characteristics of the unqualified sample, and sends the determination result and the suspected cause of the failure to the display terminal.

[0008] By adopting the above technical scheme, each module detects different key indicators of the sample, and summarizes the detection data to the control module. The control module conducts a comprehensive analysis based on the trained quality analysis model, thereby deriving a comprehensive qualified index of the sample; when the product is unqualified, the control module analyzes the suspected cause of the failure in the production link based on the characteristics of the detection data, thereby improving the efficiency and accuracy of the detection of the qualified rate of caprylic / decanoic acid triglyceride finished products.

[0009] Optionally, the system further comprises: a feed flow rate monitoring module, a stirring speed monitoring module and a temperature monitoring module; The feed flow monitoring module includes an electromagnetic flowmeter disposed in the octanoic acid feed pipeline and the glycerol feed pipeline, which is used to monitor temperature data; The stirring speed monitoring module includes a Hall sensor disposed near a rotating component of the stirring device, and is used to monitor the speed data of the stirring device; The temperature monitoring module is used to monitor the temperature data of materials during the mixing process; The feed flow rate monitoring module, the stirring speed monitoring module and the temperature monitoring module communicate with the control module; The control module receives the monitoring data sent by each monitoring module and compares it with the preset interval of the corresponding data; when at least one of the monitoring data is not in the corresponding preset interval, the control module controls the sampling device to increase the sampling frequency.

[0010] Optionally, when at least one of the monitoring data is not within the corresponding preset interval, the control module controls the sampling device to increase the sampling frequency, including: Receive the monitoring data sent by each monitoring module, compare them with the upper and lower limits of the corresponding preset interval, and calculate the absolute value of the difference between each monitoring data and the preset interval; Matching the absolute value of the difference between each monitoring data and the preset interval with the set abnormality level threshold to divide the abnormality level of each monitoring data; the abnormality level includes mild abnormality, moderate abnormality and severe abnormality; Set a time weight factor for each monitoring data, and modify the abnormality level based on the duration of the data deviation from the preset interval and the time weight factor; The corresponding sampling frequency is matched based on the abnormality level of each monitoring data.

[0011] Optionally, a time weight factor is set for each monitoring data, and the abnormality level is corrected according to the duration of the data deviation from the preset interval combined with the time weight factor, including: Set the corresponding time weight factor a for the temperature data; ; When the temperature data deviates from the corresponding preset interval, the timing module is started to record the duration t of the temperature data deviating from the preset interval; wherein the timing module is connected to the control module and transmits the recorded time information to the control module in real time; Set the time threshold interval of temperature data; the time threshold interval includes a mild time interval , Medium time interval , heavy time interval ; ; When the temperature data deviates from the preset interval for a duration t within the mild time interval, the abnormality level is maintained at the level determined by the absolute value of the difference; When the temperature data deviates from the preset interval for a duration t in the medium time interval Intra-or heavy time interval , multiply the value corresponding to the abnormality level originally divided according to the absolute value of the difference by Get the value of the modified abnormality level; when it is in the medium time interval, for ; When in the heavy time interval, for ; The abnormality level is corrected based on the corrected abnormality level value.

[0012] Optionally, the control module obtains the suspected cause of the failure in the production process according to the detection data characteristics of the unqualified sample through a fault analysis model; the generation step of the fault analysis model includes: S201. Analyze the potential causal relationship between different production factors and finished product quality indicators based on historical data, identify and record the production factors and quality indicators that appear in pairs to form a number of variable pairs; and use the production factors and quality indicators as nodes of the Bayesian network; the production factors include: feed flow rate, stirring speed, temperature; the quality indicators include: density, dielectric constant, glyceride composition; S202, using a correlation analysis method to screen out variable pairs with significant correlation for the plurality of variable pairs obtained in step S201; S203, performing a conditional independence test on the variable pairs screened out in step S202 to determine whether there is a conditional independence relationship between two variables in the variable pairs; S204, establishing directed edges for the variable pairs with conditional relationships and determining the directions of the directed edges, and obtaining a Bayesian network after all the variable pairs have completed the establishment of directed edges; S205, using the maximum likelihood estimation method based on historical data to preliminarily estimate the conditional probability distribution of each node in the Bayesian network to obtain a preliminary conditional probability distribution table; S206, constructing a likelihood function according to the nodes and directed edges in the Bayesian network structure; the likelihood function is used to reflect the probability of historical data under given initial condition probability distribution parameters; S207, using a gradient descent optimization algorithm to find the parameter value when the likelihood function reaches the maximum value and replacing the corresponding old parameter in the Bayesian network with the parameter value when the likelihood function reaches the maximum value; When the comprehensive qualified index is lower than the preset threshold, the current corresponding production factor monitoring data and sample detection data are input into the trained fault analysis model; the fault analysis model calculates the posterior probability of quality problems caused by different production factors; the fault causes are sorted according to the size of the posterior probability to determine the primary suspected cause of the fault.

[0013] Optionally, the system further comprises an alarm device; when the caprylic / capric triglyceride sample is determined to be unqualified, the control module controls the alarm device to operate.

[0014] Optionally, the temperature monitoring module includes a plurality of temperature sensors evenly arranged along the inner side of the tank wall of the stirring device.

[0015] Optionally, during the transportation process, the temperature inside the pipeline of the transportation pipeline device is maintained within a preset temperature range.

[0016] In the second aspect, the present application provides a method for detecting the qualified rate of finished products of caprylic / decanoic acid triglyceride, which adopts the following technical scheme.

[0017] A method for detecting the qualified rate of finished products of caprylic / decanoic acid triglyceride comprises: Control the sampling device to sample the finished product of caprylic / decanoic acid triglyceride on the production line; Control the delivery pipeline device to deliver the sampled caprylic and capric triglyceride sample The flow dividing device is controlled to divide the caprylic acid triglyceride sample into a first branch flow, a second branch flow and a third branch flow; the caprylic acid triglyceride sample of the first branch flow flows into the density detection module; The caprylic acid triglyceride sample of the second branch flows into the dielectric constant detection module; the caprylic acid triglyceride sample of the third branch flows into the chromatographic analysis module; A comprehensive qualification index of the sample is obtained based on the sample dielectric constant, sample density, analysis results of glyceride substances and the trained quality analysis model; when the comprehensive qualification index is lower than a preset threshold, the caprylic / capric triglyceride sample is judged to be unqualified, and the control module obtains the suspected cause of the fault in the production link based on the detection data characteristics of the unqualified sample, and sends the judgment result and the suspected cause of the fault to the display terminal. BRIEF DESCRIPTION OF THE DRAWINGS

[0018] Figure 1 This is a system block diagram of a finished product qualification rate detection system of caprylic / capric triglyceride according to an embodiment of the present application; Figure 2 This is a flow chart of a method for detecting the qualified rate of finished products of caprylic / decanoic acid triglyceride according to an embodiment of the present application; In the figure, 101 is a sampling device; 102 is a delivery pipeline device; 103 is a flow dividing device; 104 is a density detection module; 105 is a dielectric constant detection module; 106 is a chromatographic analysis module; and 107 is a control module. DETAILED DESCRIPTION

[0019] The following is combined with Figure 1-2 The present application is further described with specific embodiments: The present application embodiment discloses a finished product qualification rate detection system of caprylic / capric triglyceride, comprising: The sampling device 101 is used to sample the finished product of caprylic / decanoic acid triglyceride on the production line.

[0020] A delivery pipeline device 102 is connected to the sampling device; the delivery pipeline device is used to deliver the sampled caprylic / capric triglyceride sample; A flow dividing device 103 is connected to the delivery pipeline device; the flow dividing device divides the caprylic / capric triglyceride sample into a first branch flow, a second branch flow and a third branch flow; Density detection module 104, the caprylic acid triglyceride sample of the first branch flows into the density detection module; the density detection module is used to measure the density of the sample. Specifically, density is one of the important indicators for judging the quality of the finished caprylic acid triglyceride. The density detection module receives the first branch sample after diversion and uses corresponding physical detection principles, such as vibration principle, to measure the density of the sample.

[0021] Dielectric constant detection module 105, the caprylic acid triglyceride sample of the second branch flows into the dielectric constant detection module; the dielectric constant detection module is used to measure the dielectric constant of the sample. Specifically, for caprylic acid triglyceride, the dielectric constants of products of different qualities will be different, and the dielectric constant of the sample of the second branch is detected by the dielectric constant detection module.

[0022] Chromatographic analysis module 106, the caprylic acid triglyceride sample of the third branch flows into the chromatographic analysis module; the chromatographic analysis module is used to analyze the glyceride substances in the sample. Specifically, gas chromatography and liquid chromatography can be used to separate and determine the content of each component based on the difference in the distribution coefficient between the stationary phase and the mobile phase.

[0023] Control module 107, the control module is electrically connected to the density detection module, the dielectric constant detection module and the chromatographic analysis module; the control module obtains the comprehensive qualified index of the sample based on the sample dielectric constant, sample density, glyceride substance analysis results and the trained quality analysis model; when the comprehensive qualified index is lower than the preset threshold, the caprylic acid triglyceride sample is judged to be unqualified, and the control module obtains the suspected cause of the failure in the production link according to the detection data characteristics of the unqualified sample, and sends the judgment result and the suspected cause of the failure to the display terminal. Specifically, the display terminal can refer to a computer display screen, a dedicated industrial control display and other equipment, and the display terminal receives and displays the test result information sent by the control module, including the comprehensive qualified index of the sample and the suspected cause of the failure in the production link when the product is unqualified, so as to facilitate production, quality control and other related personnel to intuitively understand the product quality status and production problems.

[0024] In the application, since the quality of the finished product of caprylic acid triglyceride is affected by many factors, this application obtains samples from the production line through a sampling device, and then a conveying pipeline device ensures that the samples are transported to the diversion device under a stable temperature environment. After diversion, samples from different tributaries flow into the corresponding density detection module, dielectric constant detection module and chromatographic analysis module respectively. Each module detects different key indicators of the sample, and summarizes the detection data to the control module. The control module conducts a comprehensive analysis based on the trained quality analysis model, thereby deriving a comprehensive qualified index of the sample; when the product is unqualified, the control module analyzes the suspected cause of the failure in the production link based on the characteristics of the detection data, thereby improving the efficiency and accuracy of the qualified rate detection of the finished product of caprylic acid triglyceride.

[0025] As a specific implementation of a finished product qualification rate detection system for caprylic / decanoic acid triglyceride, the system further includes: a feed flow rate monitoring module, a stirring speed monitoring module and a temperature monitoring module; The feed flow monitoring module includes an electromagnetic flowmeter disposed in the octanoic acid feed pipeline and the glycerol feed pipeline, which is used to monitor temperature data; The stirring speed monitoring module includes a Hall sensor disposed near a rotating component of the stirring device, and is used to monitor the speed data of the stirring device; The temperature monitoring module is used to monitor the temperature data of materials during the mixing process; The feed flow rate monitoring module, the stirring speed monitoring module and the temperature monitoring module communicate with the control module; The control module receives the monitoring data sent by each monitoring module and compares it with the preset interval of the corresponding data; when at least one of the monitoring data is not in the corresponding preset interval, the control module controls the sampling device to increase the sampling frequency.

[0026] Specifically, the control module receives relevant data from each monitoring module in real time, and compares it with the preset interval of the corresponding data. Once at least one monitoring data exceeds its preset interval, it means that abnormal factors affecting product quality may have occurred in the production process, such as unstable feed flow rate may cause imbalance in component ratio, abnormal stirring speed may affect the uniformity of material mixing, abnormal temperature may even change the reaction process and product characteristics, etc. At this time, the control module controls the sampling device to increase the sampling frequency, so that more samples can be obtained for detection and analysis, thereby timely capturing potential quality changes in the production process and discovering product quality problems in advance before they occur on a large scale.

[0027] As a specific implementation of a finished product qualification rate detection system for caprylic / decanoic acid triglyceride, when at least one of the monitoring data is not in the corresponding preset interval, the control module controls the sampling device to increase the sampling frequency, including: Receive the monitoring data sent by each monitoring module, compare them with the upper and lower limits of the corresponding preset interval, and calculate the absolute value of the difference between each monitoring data and the preset interval; Matching the absolute value of the difference between each monitoring data and the preset interval with the set abnormality level threshold to divide the abnormality level of each monitoring data; the abnormality level includes mild abnormality, moderate abnormality and severe abnormality; Set a time weight factor for each monitoring data, and modify the abnormality level based on the duration of the data deviation from the preset interval and the time weight factor; The corresponding sampling frequency is matched based on the abnormality level of each monitoring data.

[0028] Specifically, the control module receives the monitoring data sent by each monitoring module, and compares it with the upper and lower limits of the corresponding preset interval, calculates the absolute value of the difference between each monitoring data and the preset interval, and quantifies the degree of deviation of each monitoring data from the normal range; the absolute value of the difference between each monitoring data and the preset interval is matched with the set abnormality level threshold to divide the abnormality level into mild abnormality, moderate abnormality and severe abnormality, and distinguish different degrees of abnormal conditions. Set a time weight factor for each monitoring data, and then correct the abnormality level according to the duration of the data deviation from the preset interval combined with the time weight factor. Some abnormalities may only appear briefly and have little impact on product quality, while abnormalities that last longer are likely to have a more serious impact on product quality; moderately increase the sampling frequency when there is a mild abnormality; increase the sampling frequency when there is a moderate abnormality; further increase the sampling frequency when there is a severe abnormality, effectively avoiding the output of a large number of unqualified products due to the failure to fully detect the abnormal situation to ensure the qualified rate of the finished product of caprylic decanoic acid triglyceride.

[0029] As a specific implementation of a finished product qualification rate detection system for caprylic / decanoic acid triglyceride, a time weight factor is set for each monitoring data, and the abnormality level is corrected in combination with the time weight factor according to the duration of the data deviation from the preset interval, including: Set the corresponding time weight factor a for the temperature data; ; When the temperature data deviates from the corresponding preset interval, the timing module is started to record the duration t of the temperature data deviating from the preset interval; wherein the timing module is connected to the control module and transmits the recorded time information to the control module in real time; Set the time threshold interval of temperature data; the time threshold interval includes a mild time interval , Medium time interval , heavy time interval ; ; When the temperature data deviates from the preset interval for a duration t within the mild time interval, the abnormality level is maintained at the level determined by the absolute value of the difference; When the temperature data deviates from the preset interval for a duration t in the medium time interval Intra-or heavy time interval , multiply the value corresponding to the abnormality level originally divided according to the absolute value of the difference by Get the value of the modified abnormality level; when it is in the medium time interval, for ; When in the heavy time interval, for ; The abnormality level is corrected based on the corrected abnormality level value.

[0030] Specifically, taking temperature data abnormality as an example, a corresponding time weight factor a is set for the temperature data. When the temperature data deviates from the corresponding preset interval, the timing module will start recording the time t of its continuous deviation. By setting the temperature data time threshold interval including the mild time interval, the moderate time interval and the severe time interval, the severity of the abnormality is further measured according to the different intervals where the temperature data deviates from the preset interval duration t. When the time t is in the mild time interval, the abnormality level maintains the level determined according to the absolute value of the difference. This is because the short-term temperature fluctuation may be just an accidental situation, and the possibility of having a substantial impact on the product quality is relatively small, so the initial determination level can reasonably reflect its abnormality. When the temperature data deviates from the preset interval duration t in the moderate time interval or the severe time interval, the value corresponding to the abnormality level originally divided according to the absolute value of the difference is multiplied by the corresponding coefficient to obtain the corrected abnormality level value, so as to take into account that the long-term temperature abnormality will inevitably have a more serious impact on the production process of caprylic decanoic acid triglyceride, such as chemical reaction imbalance, material property change, etc., which greatly affects the product quality. Through this correction method that is closely integrated with the time factor, the abnormality level can be more in line with the actual situation of the impact of temperature abnormalities on product quality in actual production, so that the control module's judgment of temperature abnormalities in the production process is no longer limited to instantaneous deviations. It comprehensively considers the impact of the time dimension and reduces problems such as the decline in product qualification rate caused by temperature abnormalities.

[0031] As one implementation of a finished product qualification rate detection system for caprylic / decanoic acid triglyceride, a control module obtains suspected causes of failures in the production process according to the detection data characteristics of unqualified samples through a fault analysis model; the generation steps of the fault analysis model include: S201. Analyze the potential causal relationship between different production factors and finished product quality indicators based on historical data, identify and record the production factors and quality indicators that appear in pairs to form a number of variable pairs; and use the production factors and quality indicators as nodes of the Bayesian network; the production factors include: feed flow rate, stirring speed, temperature; the quality indicators include: density, dielectric constant, glyceride composition; S202, using a correlation analysis method to screen out variable pairs with significant correlation for the plurality of variable pairs obtained in step S201; S203, performing a conditional independence test on the variable pairs screened out in step S202 to determine whether there is a conditional independence relationship between two variables in the variable pairs; S204, establishing directed edges for the variable pairs with conditional relationships and determining the directions of the directed edges, and obtaining a Bayesian network after all the variable pairs have completed the establishment of directed edges; S205, using the maximum likelihood estimation method based on historical data to preliminarily estimate the conditional probability distribution of each node in the Bayesian network to obtain a preliminary conditional probability distribution table; S206, constructing a likelihood function according to the nodes and directed edges in the Bayesian network structure; the likelihood function is used to reflect the probability of historical data under given initial condition probability distribution parameters; S207, using a gradient descent optimization algorithm to find the parameter value when the likelihood function reaches the maximum value and replacing the corresponding old parameter in the Bayesian network with the parameter value when the likelihood function reaches the maximum value; When the comprehensive qualified index is lower than the preset threshold, the current corresponding production factor monitoring data and sample detection data are input into the trained fault analysis model; the fault analysis model calculates the posterior probability of quality problems caused by different production factors; the fault causes are sorted according to the size of the posterior probability to determine the primary suspected cause of the fault.

[0032] Specifically, based on a large amount of historical data, the potential causal relationship between different production factors (such as feed flow, stirring speed, temperature) and finished product quality indicators (density, dielectric constant, glyceride composition) is analyzed, and they are organized into variable pairs, which are then used as nodes of the Bayesian network. The correlation analysis method is used to screen out variable pairs with significant correlation, and those with weak correlation are eliminated, so that the subsequent analysis focuses on the correlation of factors that have an important impact on product quality, avoiding interference from irrelevant information and improving the accuracy of the analysis; a conditional independence test is performed to determine whether there is a conditional independence relationship between the two variables in the variable pair, and a directed edge is established and the direction is determined for the variable pairs with conditional relationships, completing the construction of the Bayesian network. Based on historical data, the maximum likelihood estimation method is used to preliminarily estimate the conditional probability distribution of each node to obtain the corresponding table, and the likelihood function is constructed according to the network structure, and the gradient descent optimization algorithm is used to find the optimal parameter value to replace the old parameter, so that the constructed fault analysis model can fit the actual production situation to the greatest extent and accurately reflect the probability relationship between each factor and quality. When the comprehensive qualified index is lower than the preset threshold, indicating that the product is unqualified, the corresponding production factor monitoring data and sample detection data are input into the trained fault analysis model. The model can calculate the posterior probability of quality problems caused by different production factors based on the logical and probability relationship constructed in the early stage, and sort the causes of the fault according to the probability, so as to determine the primary suspected cause of the fault, improve the efficiency of fault detection, and ensure the stability of the production process of caprylic / decanoic acid triglyceride and the qualified rate of the finished product.

[0033] As one embodiment of a finished product qualification rate detection system for caprylic / decanoic acid triglyceride, the system also includes an alarm device; when the caprylic / decanoic acid triglyceride sample is determined to be unqualified, the control module controls the alarm device to operate.

[0034] As one embodiment of a finished product qualification rate detection system for caprylic / decanoic acid triglyceride, a temperature monitoring module includes a plurality of temperature sensors evenly arranged along the inner side of the tank wall of the stirring device.

[0035] As one embodiment of a finished product qualification rate detection system for caprylic / decanoic acid triglyceride, during the transportation process, the temperature inside the pipeline of the transportation pipeline device is maintained in a preset temperature range.

[0036] The present application also provides a method for detecting the qualified rate of finished products of caprylic / decanoic acid triglyceride, based on the above system, comprising: S301, control the sampling device to sample the finished product of caprylic / capric triglyceride on the production line; S302, control the delivery pipeline device to deliver the sampled caprylic / capric triglyceride sample S303, controlling the flow splitting device to split the caprylic acid triglyceride sample into a first branch, a second branch and a third branch; the caprylic acid triglyceride sample of the first branch flows into a density detection module; the caprylic acid triglyceride sample of the second branch flows into a dielectric constant detection module; the caprylic acid triglyceride sample of the third branch flows into a chromatographic analysis module; S304. Obtain a comprehensive qualified index of the sample based on the sample dielectric constant, sample density, glyceride substance analysis results and the trained quality analysis model; when the comprehensive qualified index is lower than a preset threshold, the caprylic / capric triglyceride sample is determined to be unqualified, and the control module obtains the suspected cause of the failure in the production link based on the detection data characteristics of the unqualified sample, and sends the determination result and the suspected cause of the failure to the display terminal.

[0037] It should be noted that the above embodiments are only used to illustrate the present application and are not intended to limit the technical solutions described in the present application. Although the present application has been described in detail in this specification with reference to the above embodiments, a person of ordinary skill in the art should understand that a person of ordinary skill in the art can still modify or make equivalent substitutions to the present application, and all technical solutions and improvements thereof that do not depart from the spirit and scope of the present application should be included in the scope of the claims of the present application.

Claims

1. A finished product qualification rate detection system for caprylic / decanoic acid triglyceride, characterized in that: include: A sampling device, used for sampling the finished product of caprylic / decanoic acid triglyceride on the production line; A delivery pipeline device connected to the sampling device; the delivery pipeline device is used to deliver the sampled caprylic / capric triglyceride sample; A flow dividing device is connected to the delivery pipeline device; the flow dividing device divides the caprylic / capric triglyceride sample into a first branch flow, a second branch flow and a third branch flow; Density detection module, the caprylic / capric triglyceride sample of the first branch flows into the density detection module; the density detection module is used to measure the density of the sample; A dielectric constant detection module, the caprylic / capric triglyceride sample of the second branch flows into the dielectric constant detection module; the dielectric constant detection module is used to measure the dielectric constant of the sample; Chromatographic analysis module, the caprylic / capric triglyceride sample of the third branch flows into the chromatographic analysis module; the chromatographic analysis module is used to analyze the glyceride substances in the sample; A control module, wherein the control module is electrically connected to the density detection module, the dielectric constant detection module and the chromatographic analysis module; the control module obtains a comprehensive qualification index of the sample based on the sample dielectric constant, the sample density, the analysis results of the glyceride substances and the trained quality analysis model; when the comprehensive qualification index is lower than a preset threshold, the caprylic / capric triglyceride sample is determined to be unqualified, and the control module obtains the suspected cause of the failure in the production link according to the detection data characteristics of the unqualified sample, and sends the determination result and the suspected cause of the failure to the display terminal.

2. The finished product qualification rate detection system of caprylic / decanoic acid triglyceride according to claim 1, characterized in that: The system also includes: a feed flow monitoring module, a stirring speed monitoring module and a temperature monitoring module; The feed flow monitoring module includes an electromagnetic flowmeter disposed in the octanoic acid feed pipeline and the glycerol feed pipeline, which is used to monitor temperature data; The stirring speed monitoring module includes a Hall sensor disposed near a rotating component of the stirring device, and is used to monitor the speed data of the stirring device; The temperature monitoring module is used to monitor the temperature data of materials during the mixing process; The feed flow rate monitoring module, the stirring speed monitoring module and the temperature monitoring module communicate with the control module; The control module receives the monitoring data sent by each monitoring module and compares it with the preset interval of the corresponding data; when at least one of the monitoring data is not in the corresponding preset interval, the control module controls the sampling device to increase the sampling frequency.

3. The finished product qualification rate detection system of caprylic / decanoic acid triglyceride according to claim 2, characterized in that: When at least one of the monitoring data is not within the corresponding preset interval, the control module controls the sampling device to increase the sampling frequency, including: Receive the monitoring data sent by each monitoring module, compare them with the upper and lower limits of the corresponding preset interval, and calculate the absolute value of the difference between each monitoring data and the preset interval; Matching the absolute value of the difference between each monitoring data and the preset interval with the set abnormality level threshold to divide the abnormality level of each monitoring data; the abnormality level includes mild abnormality, moderate abnormality and severe abnormality; Set a time weight factor for each monitoring data, and modify the abnormality level based on the duration of the data deviation from the preset interval and the time weight factor; The corresponding sampling frequency is matched based on the abnormality level of each monitoring data.

4. The finished product qualification rate detection system of caprylic / decanoic acid triglyceride according to claim 3, characterized in that: Set a time weight factor for each monitoring data, and modify the abnormality level based on the duration of the data deviation from the preset interval, combined with the time weight factor, including: Set the corresponding time weight factor a for the temperature data; ; When the temperature data deviates from the corresponding preset interval, the timing module is started to record the duration t of the temperature data deviating from the preset interval; wherein the timing module is connected to the control module and transmits the recorded time information to the control module in real time; Set the time threshold interval of temperature data; the time threshold interval includes a mild time interval , Medium time interval , heavy time interval ; ; When the temperature data deviates from the preset interval for a duration t within the mild time interval, the abnormality level is maintained at the level determined by the absolute value of the difference; When the temperature data deviates from the preset interval for a duration t in the medium time interval Intra-or heavy time interval , multiply the value corresponding to the abnormality level originally divided according to the absolute value of the difference by Get the value of the modified abnormality level; when it is in the medium time interval, for ; When in the heavy time interval, for ; The abnormality level is corrected based on the corrected abnormality level value.

5. A finished product qualification rate detection system for caprylic / decanoic acid triglyceride according to claim 4, characterized in that: The control module obtains the suspected cause of the failure in the production process according to the detection data characteristics of the unqualified samples through a fault analysis model; the generation steps of the fault analysis model include: S201. Analyze the potential causal relationship between different production factors and finished product quality indicators based on historical data, identify and record the production factors and quality indicators that appear in pairs to form a number of variable pairs; and use the production factors and quality indicators as nodes of the Bayesian network; the production factors include: feed flow rate, stirring speed, temperature; the quality indicators include: density, dielectric constant, glyceride composition; S202, using a correlation analysis method to screen out variable pairs with significant correlation for the plurality of variable pairs obtained in step S201; S203, performing a conditional independence test on the variable pairs screened out in step S202 to determine whether there is a conditional independence relationship between two variables in the variable pairs; S204, establishing directed edges for the variable pairs with conditional relationships and determining the directions of the directed edges, and obtaining a Bayesian network after all the variable pairs have completed the establishment of directed edges; S205, using the maximum likelihood estimation method based on historical data to preliminarily estimate the conditional probability distribution of each node in the Bayesian network to obtain a preliminary conditional probability distribution table; S206, constructing a likelihood function according to the nodes and directed edges in the Bayesian network structure; the likelihood function is used to reflect the probability of historical data under given initial condition probability distribution parameters; S207, using a gradient descent optimization algorithm to find the parameter value when the likelihood function reaches the maximum value and replacing the corresponding old parameter in the Bayesian network with the parameter value when the likelihood function reaches the maximum value; When the comprehensive qualified index is lower than the preset threshold, the current corresponding production factor monitoring data and sample detection data are input into the trained fault analysis model; the fault analysis model calculates the posterior probability of quality problems caused by different production factors; the fault causes are sorted according to the size of the posterior probability to determine the primary suspected cause of the fault.

6. A finished product qualification rate detection system for caprylic / decanoic acid triglyceride according to claim 5, characterized in that: The system also includes an alarm device; when the caprylic / capric triglyceride sample is determined to be unqualified, the control module controls the alarm device to operate.

7. A finished product qualification rate detection system for caprylic / decanoic acid triglyceride according to claim 6, characterized in that: The temperature monitoring module comprises a plurality of temperature sensors evenly arranged along the inner side of the tank wall of the stirring device.

8. The finished product qualification rate detection system of caprylic / decanoic acid triglyceride according to claim 6, characterized in that: During the transportation process, the temperature inside the pipeline of the transportation pipeline device is maintained within a preset temperature range.

9. A method for detecting the qualified rate of finished products of caprylic / decanoic acid triglyceride, based on the system according to any one of claims 1 to 8, characterized in that: include: Control the sampling device to sample the finished product of caprylic / decanoic acid triglyceride on the production line; Control the delivery pipeline device to deliver the sampled caprylic and capric triglyceride sample The flow dividing device is controlled to divide the caprylic acid triglyceride sample into a first branch flow, a second branch flow and a third branch flow; the caprylic acid triglyceride sample of the first branch flow flows into the density detection module; The caprylic acid triglyceride sample of the second branch flows into the dielectric constant detection module; the caprylic acid triglyceride sample of the third branch flows into the chromatographic analysis module; A comprehensive qualification index of the sample is obtained based on the sample dielectric constant, sample density, analysis results of glyceride substances and the trained quality analysis model; when the comprehensive qualification index is lower than a preset threshold, the caprylic / capric triglyceride sample is judged to be unqualified, and the control module obtains the suspected cause of the fault in the production link based on the detection data characteristics of the unqualified sample, and sends the judgment result and the suspected cause of the fault to the display terminal.

Images