Methyl methacrylate production environment data monitoring method

Abstract

The invention relates to the technical field of data processing, in particular to a methyl methacrylate production environment data monitoring method. Obtaining an abnormal change point according to the data difference characteristics of the historical temperature time sequence; obtaining a reaction deposition characteristic value according to the temperature difference characteristic of the adjacent abnormal change points; judging whether the abnormal change point is a deposition feature point or not according to the reaction deposition feature value; obtaining a temperature fluctuation sequence according to all the deposition feature points; and obtaining a deposition change degree value according to similar characteristics of the temperature fluctuation sequence and a preset ascending sequence. According to the deposition change degree value and the reaction deposition characteristic value, the self-adaptive lag coefficient weight of the abnormal change point is obtained; the temperature is predicted according to the ARIMA prediction model containing the self-adaptive lag coefficient weight; and the production process is monitored according to the predicted temperature, so that the influence of a solid catalyst deposition phenomenon on temperature monitoring is avoided, and the accuracy of temperature monitoring is improved.

Description

Technical Field

[0001] The invention relates to the technical field of data processing, and in particular to a method for monitoring methyl methacrylate production environment data. Background Art

[0002] Methyl methacrylate is an important organic chemical raw material with a wide range of uses. The main method for industrial synthesis of methyl methacrylate is the process of using methyl acetate and formaldehyde as raw materials to produce methyl acrylate, methyl propionate and methyl methacrylate through condensation-hydrogenation-condensation. In the industrial production process, the temperature of the reaction process has a great influence on the quality of the synthesis result. Therefore, it is necessary to monitor the temperature in the reactor during the production process to ensure the quality of methyl methacrylate.

[0003] Since the production of methyl methacrylate takes a long time, the temperature value of the future period can be predicted based on the temperature changes in the historical reaction process. The existing ARIMA time series model can predict the temperature data of the future period through the historical temperature data; and judge whether there is an abnormal trend in the temperature during the production process based on the difference between the predicted temperature and the actual temperature. However, the solid catalyst in the production process of methyl methacrylate is prone to normal deposition and aggregation, resulting in intense heat release in local areas, causing temperature fluctuations; the uncertainty of the deposition and aggregation of the solid catalyst is likely to cause the ARIMA prediction model to deviate from the temperature prediction, resulting in a large difference between the predicted temperature and the actual temperature, which ultimately leads to low accuracy in monitoring temperature anomalies during the production of methyl methacrylate. Summary of the invention

[0004] In order to solve the technical problem that the normal deposition and aggregation of solid catalyst in the above-mentioned methyl methacrylate production process causes temperature fluctuations and affects the accuracy of temperature abnormality monitoring, the purpose of the present invention is to provide a methyl methacrylate production environment data monitoring method, and the technical scheme adopted is as follows: Get the historical temperature time series of the reactor at the current moment in the production process of methyl methacrylate; Obtain an abnormal change point according to the data difference characteristics between the data point and the adjacent data points in the historical temperature time series, and the difference characteristics between the data point and the overall data change trend; obtain a reaction deposition characteristic value according to the temperature difference characteristics of the adjacent abnormal change points; According to the reaction deposition characteristic value, determine whether the abnormal change point is a deposition characteristic point; if there is a deposition characteristic point, construct a temperature change curve according to all deposition characteristic points, and obtain a temperature fluctuation sequence according to the temperature difference characteristics of the peaks and valleys in the temperature change curve; obtain a deposition change degree value according to the similarity characteristics of the temperature fluctuation sequence and the data change of the preset ascending sequence; The original lag coefficient weight of the abnormal change point in the ARIMA prediction model is adaptively adjusted according to the deposition change degree value and the reaction deposition characteristic value to obtain an adaptive lag coefficient weight; the temperature at a future moment is predicted according to the ARIMA prediction model including the adaptive lag coefficient weight to obtain a predicted temperature; and the production process is monitored according to the predicted temperature.

[0005] Furthermore, the step of obtaining abnormal change points according to the data difference characteristics between the data point and adjacent data points in the historical temperature time series and the difference characteristics between the data point and the overall data change trend includes: Constructing a temperature-time coordinate system of the historical temperature time series, calculating the sum of the Euclidean distances between any data point in the historical temperature time series and the adjacent data points before and after in the temperature-time coordinate system, and obtaining a neighborhood difference characteristic value of the arbitrary data point; performing time series decomposition on the historical temperature time series in the temperature-time coordinate system, obtaining a temperature change trend curve, calculating and normalizing the Euclidean distance between the arbitrary data point and the data point at the same time in the temperature change trend curve, and obtaining a trend difference characteristic value of the arbitrary data point; calculating and normalizing the inverse of the neighborhood difference characteristic value, and obtaining a neighborhood similarity; calculating the average value of the trend difference characteristic value and the neighborhood similarity, and obtaining the degree of abnormality of the arbitrary data point; When the abnormal degree of the arbitrary data point exceeds a preset abnormal threshold, the arbitrary data point is an abnormal change point.

[0006] Furthermore, the step of obtaining the reaction deposition characteristic value according to the temperature difference characteristics of adjacent abnormal change points includes: Calculate the temperature difference between any abnormal change point and the next adjacent abnormal change point to obtain the temperature difference; calculate the inverse of the absolute value of the sum of all temperature differences to obtain the temperature stability characteristic value; calculate the ratio of the number of abnormal change points to a preset constant to obtain the quantity characteristic value; calculate the product of the quantity characteristic value and the temperature stability characteristic value to obtain the reaction deposition characteristic value.

[0007] Further, the step of judging whether the abnormal change point is a deposition characteristic point according to the reaction deposition characteristic value includes: When the reaction deposition characteristic value exceeds a preset deposition threshold, the abnormal change point is a deposition characteristic point.

[0008] Furthermore, the step of obtaining a temperature fluctuation sequence according to the temperature difference characteristics of the peaks and valleys in the temperature change curve includes: The longitudinal vertical distance between any peak point in the temperature change curve and the nearest trough point in the future time is calculated to obtain the amplitude characteristic value; all the amplitude characteristic values ​​are sorted in chronological order to obtain the temperature fluctuation sequence.

[0009] Furthermore, the step of obtaining the deposition change degree value according to the similar characteristics of the data change of the temperature fluctuation sequence and the preset ascending sequence includes: The Spearman correlation coefficient between the temperature fluctuation sequence and the preset ascending sequence is calculated and positively correlated to obtain the deposition change degree value.

[0010] Furthermore, the step of adaptively adjusting the original lag coefficient weight of the abnormal change point in the ARIMA prediction model according to the deposition change degree value and the reaction deposition characteristic value to obtain the adaptive lag coefficient weight includes: When the reaction deposition characteristic value does not exceed the preset deposition threshold, the adaptive hysteresis coefficient weight of the abnormal change point is a constant of 0; when the deposition change degree value does not exceed the preset change threshold, the product of the original hysteresis coefficient weight of the abnormal data point and the deposition change degree value is calculated to obtain the adaptive hysteresis coefficient weight; when the deposition change degree value exceeds the preset change threshold, the sum of the deposition change degree value and a preset second constant is calculated to obtain the adjustment amount; the product of the adjustment amount and the original hysteresis coefficient weight is calculated to obtain the adaptive hysteresis coefficient weight.

[0011] Furthermore, the step of monitoring the production process according to the predicted temperature includes: The absolute value of the difference between the predicted temperature and the actual measured temperature at the same time is calculated to obtain a temperature error value; when the temperature error value exceeds a preset error threshold, an abnormality occurs in the methyl methacrylate production process.

[0012] The present invention has the following beneficial effects: In the present invention, obtaining abnormal change points can determine whether there are data points suspected of abnormal temperature trends in the historical temperature time series, and then adaptively adjust the weights of abnormal change points in the prediction process; since the abnormal temperature trend in the production process may be caused by the deposition of solid catalysts and the unstable heating of the reactor, obtaining the reaction deposition characteristic value can characterize the cause of the abnormal change point, and further improve the accuracy of the abnormal change point weight setting and temperature anomaly monitoring. Since there is uncertainty in the deposition phenomenon of solid catalysts in the reaction process, the changing trends of different deposition phenomena will lead to differences in the temperature change trends, so obtaining the temperature fluctuation sequence can characterize the changing trend of the deposition phenomenon; obtaining the deposition change degree value can quantify the changing trend of the deposition phenomenon, and further improve the accuracy of the abnormal change point weight setting. Obtaining the adaptive hysteresis coefficient weight can change the importance of the abnormal change point in the prediction process, so that the temperature prediction results caused by different factors can be different, and finally monitoring according to the predicted temperature can avoid the influence of the solid catalyst deposition phenomenon on temperature monitoring, and improve the accuracy of temperature monitoring. BRIEF DESCRIPTION OF THE DRAWINGS

[0013] In order to more clearly illustrate the technical solutions and advantages in the embodiments of the present invention or the prior art, the drawings required for use in the embodiments or the prior art descriptions are briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0014] Figure 1 A flow chart of a method for monitoring methyl methacrylate production environment data provided by one embodiment of the present invention. DETAILED DESCRIPTION

[0015] In order to further explain the technical means and effects adopted by the present invention to achieve the predetermined invention purpose, the following is a detailed description of a methyl methacrylate production environment data monitoring method proposed by the present invention, its specific implementation, structure, characteristics and effects, in conjunction with the accompanying drawings and preferred embodiments. In the following description, different "one embodiment" or "another embodiment" does not necessarily refer to the same embodiment. In addition, specific features, structures or characteristics in one or more embodiments may be combined in any suitable form.

[0016] Unless defined otherwise, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention belongs.

[0017] The specific scheme of the method for monitoring methyl methacrylate production environment data provided by the present invention is described in detail below with reference to the accompanying drawings.

[0018] See also Figure 1 , which shows a flow chart of a method for monitoring methyl methacrylate production environment data provided by an embodiment of the present invention, the method comprising the following steps: Step S1, obtaining a historical temperature time series of a reactor in a methyl methacrylate production process at a current moment.

[0019] The main method for industrial production of methyl methacrylate is to use methyl acetate and formaldehyde as raw materials and produce methyl acrylate, methyl propionate and methyl methacrylate through condensation-hydrogenation-condensation. Solid catalysts such as silicon dioxide are required for catalytic action in the condensation process. The temperature in the reactor during the condensation reaction has an important influence on the reaction result. Therefore, the historical temperature time series of the reactor at the current moment in the production process of methyl methacrylate is obtained, and the temperature is collected by the thermocouple temperature sensor in the reactor; in the embodiment of the present invention, the temperature data of 5 minutes before the current moment in the condensation process is collected as the historical temperature time series at the current moment, and the collection frequency is once per second; with the update of the current moment, the corresponding historical temperature time series will be updated synchronously; the implementer can determine the collection duration and collection frequency by himself.

[0020] Step S2, obtaining an abnormal change point according to the data difference characteristics between the data point and the adjacent data points in the historical temperature time series, and the difference characteristics between the data point and the overall data change trend; obtaining a reaction deposition characteristic value according to the temperature difference characteristics of the adjacent abnormal change points.

[0021] During the condensation reaction, temperature control is crucial. The temperature changes at adjacent moments during the reaction process are regular, so the predicted temperature based on historical normal temperature data is similar to the actual temperature at the same moment. When an abnormal temperature trend occurs, there will be a difference between the predicted result based on normal temperature data and the actual measurement result, and then it can be judged whether the condensation process is abnormal based on the degree of difference. Since solid catalysts may deposit and aggregate during the condensation reaction, resulting in intense heat release in local areas of the reactor, the temperature shows an abnormal trend. However, the deposition of solid catalysts is a normal situation during the reaction process, and as the reaction proceeds, the temperature will tend to normal. Therefore, in the temperature prediction process, it is necessary to reduce the uncertain influence of the deposition and aggregation of solid catalysts and improve the accuracy of temperature monitoring during the condensation reaction.

[0022] In the normal condensation reaction process, when the solid catalyst is evenly distributed in the reactor, the temperature monitoring data in the early stage of the reaction shows a stable trend with a small degree of fluctuation; in the reaction stage, due to the effect of the solid catalyst, the temperature will show a slow upward trend. Therefore, in the normal reaction process, the temperature monitored at any time is consistent with the overall temperature trend change. If there is a large difference between the temperature monitored at any time and the overall temperature trend, an abnormal temperature trend may appear at that moment. Furthermore, individual noise points may appear in the acquisition process, and the noise points will also have a large difference with the overall temperature trend change, but the noise points will also have a large difference with the temperature data of the adjacent moments before and after, and the real temperature abnormal trend data points are correlated. Therefore, the abnormal change points can be obtained based on the data difference characteristics between the data points and the adjacent data points in the historical temperature time series, and the difference characteristics between the data points and the overall data change trend.

[0023] Preferably, in an embodiment of the present invention, the step of obtaining an abnormal change point includes: constructing a temperature-time coordinate system of a historical temperature time series, calculating the sum of the Euclidean distances of any data point in the historical temperature time series and the adjacent data points before and after in the temperature-time coordinate system, and obtaining a neighborhood difference characteristic value of the arbitrary data point; when the neighborhood difference characteristic value is larger, it means that the arbitrary data point is more likely to be a noise point, and the temperature data before and after the moment are less related. In the temperature-time coordinate system, the historical temperature time series is decomposed in time series to obtain a temperature change trend curve. It should be noted that the time series decomposition of the sequence belongs to the prior art, and the specific steps are not repeated. The temperature change trend curve reflects the overall change trend of the temperature time series. Calculate the Euclidean distance of any data point and the data point at the same time in the temperature change trend curve and normalize it to obtain the trend difference characteristic value of the arbitrary data point; when the trend difference characteristic value is larger, it means that the temperature of the arbitrary data point is more different from the overall temperature change trend, and it is more likely to represent the temperature abnormal trend and noise point. Calculate the inverse of the neighborhood difference characteristic value and normalize it to obtain the neighborhood similarity; when the neighborhood similarity is larger, it means that the arbitrary data point is less likely to be a noise point. Calculate the average value of the trend difference characteristic value and the neighborhood similarity to obtain the degree of abnormality of the arbitrary data point; the greater the degree of abnormality of the arbitrary data point, the more likely it is that the arbitrary data point represents an abnormal temperature trend. When the degree of abnormality of the arbitrary data point exceeds the preset abnormal threshold, the arbitrary data point is an abnormal change point; in an embodiment of the present invention, the preset abnormal threshold is 0.6, and the preset abnormal threshold can better distinguish the abnormal change points of suspected temperature trend abnormalities, which can be determined by the implementer according to the implementation scenario; the abnormal change point represents the existence of a suspected abnormal temperature trend in the condensation reaction process at this moment. The formula for obtaining the degree of abnormality includes: In the formula, Q represents the abnormality of the arbitrary data point, and D represents the Euclidean distance between the arbitrary data point and the data point at the same time in the temperature change trend curve. represents normalization, Indicates the trend difference characteristic value; represents the Euclidean distance between any data point and the previous adjacent data point. It represents the Euclidean distance between any data point and the next adjacent data point. represents the neighborhood difference eigenvalue; Represents neighborhood similarity.

[0024] Furthermore, the reason for the abnormal change point may be that the solid catalyst deposits and aggregates during the stirring process, resulting in intense heat release in the local area; it may also be caused by unstable heating of the reactor, such as continuous heating causing the temperature to rise rapidly, or interruption of heating causing the temperature to drop rapidly, causing the temperature to deviate from the normal temperature change trend. As the deposition and aggregation of the solid catalyst will gradually dissociate as the reaction proceeds, and the temperature will tend to the normal level, the temperature change caused by the deposition process needs to be considered in the prediction process, so that the temperature prediction value is closer to the temperature level caused by the deposition situation to avoid false alarms; and for the temperature anomaly caused by unstable heating of the reactor, such abnormal data points need to be discarded in the prediction process, so that the prediction result is closer to the normal temperature level, and then the difference between the predicted value and the actual abnormal temperature value is different, and an early warning is issued. Therefore, it is necessary to first determine whether the abnormal change point is caused by the deposition of the solid catalyst. The unstable heating of the reactor will radiate to the entire internal area, and the resulting temperature offset will continue to change synchronously, such as continuous temperature increase deviation, continuous dimensionality reduction deviation trend item, and the overall abnormal data points will show the characteristics of gradually rising or falling temperature. For the temperature rise caused by the solid catalyst, the temperature rises in some areas of the reactor, and the high temperature is transferred to other areas due to the heat of the solution stirring, causing the temperature to gradually drop; as part of the solid catalyst is deposited again, the temperature will fluctuate again, so the solid catalyst will cause the temperature to fluctuate rather than continuously change in one direction. Therefore, the reaction deposition characteristic value is obtained based on the temperature difference characteristics of adjacent abnormal change points.

[0025] Preferably, in an embodiment of the present invention, the step of obtaining the reaction deposition characteristic value includes: calculating the temperature difference between any abnormal change point and the next adjacent abnormal change point to obtain the temperature difference; calculating the inverse of the absolute value of the sum of all temperature differences to obtain the temperature stability characteristic value; when the temperature stability characteristic value is larger, it means that the absolute value of the sum of all temperature differences is smaller, and the temperature of the abnormal change point is more likely to fluctuate; when the temperature stability characteristic value is smaller, it means that the temperature of the abnormal change point is more likely to be a single change trend, and it is more likely to be caused by unstable heating of the reactor. Calculate the ratio of the number of abnormal change points to a preset constant to obtain a quantitative characteristic value; in the embodiment of the present invention, the preset constant is 5, the purpose is to avoid the analysis error of the temperature stability characteristic value caused by the small number of abnormal change points; the implementer can determine it by himself according to the implementation scenario, the larger the quantitative characteristic value, the more the number of abnormal change points, and the greater the credibility of the temperature stability characteristic value. Calculate the product of the quantitative characteristic value and the temperature stability characteristic value to obtain the reaction deposition characteristic value; when the reaction deposition characteristic value is larger, it means that the abnormal change point is more likely to be caused by solid catalyst deposition. The formula for obtaining the reaction deposition characteristic value includes:

[0026] In the formula, F represents the characteristic value of sedimentation, N represents the number of abnormal change points, represents the temperature of the nth abnormal change point, Indicates The temperature of the abnormal change point, Represents the temperature difference, represents the temperature stability characteristic value, a represents the preset constant, Represents a quantitative feature value.

[0027] Step S3, judging whether the abnormal change point is a deposition characteristic point according to the reaction deposition characteristic value; if there is a deposition characteristic point, constructing a temperature change curve according to all deposition characteristic points, and obtaining a temperature fluctuation sequence according to the temperature difference characteristics of the peaks and valleys in the temperature change curve; obtaining a deposition change degree value according to the similar characteristics of the data changes of the temperature fluctuation sequence and the preset ascending sequence.

[0028] When the reaction deposition characteristic value is smaller, it means that the suspected temperature trend abnormality is more likely to be caused by the instability of the reactor; therefore, the abnormal change point is judged as a deposition characteristic point according to the reaction deposition characteristic value, specifically including: when the reaction deposition characteristic value exceeds the preset deposition threshold, the abnormal change point is a deposition characteristic point; in the embodiment of the present invention, the preset deposition threshold is 0.3, which can be determined by the implementer according to the implementation scenario. The deposition characteristic point characterizes the suspected temperature trend abnormality caused by the solid catalyst deposition in the historical temperature time series. For the temperature trend abnormality caused by the solid catalyst deposition, it is necessary to further analyze the possible temperature change characteristics so that the prediction results are more in line with the actual temperature and avoid temperature misreporting caused by the deposition situation. Under the continuous stirring of the reactor at different times, the catalyst deposition aggregation may further expand or gradually dissociate; in the temperature fluctuation process of the deposition characteristic point, if the amplitude of the fluctuation gradually decreases, it means that the temperature rise caused by the solid catalyst deposition is gradually reduced, and the deposited and aggregated solid catalyst is gradually dissociating; and then the deposition and aggregation part has no effect on the temperature monitoring after dissociation, and the weight of the deposition characteristic point needs to be reduced in the prediction process so that the temperature data in the prediction process is more normal. When the temperature fluctuation amplitude of the deposition characteristic point gradually increases, it means that the deposition aggregation has not been dissociated and more solid catalysts have been aggregated, and then the deposition phenomenon will affect the temperature at the future time. Therefore, it is necessary to provide the weight of the deposition characteristic point in the prediction process so that the prediction result can better reflect the deposition characteristics and avoid a large deviation between the predicted temperature and the actual temperature and false alarm. Therefore, if there are deposition characteristic points, a temperature change curve is constructed according to all deposition characteristic points, and a temperature fluctuation sequence is obtained according to the temperature difference characteristics of the peaks and valleys in the temperature change curve; preferably, in an embodiment of the present invention, the step of obtaining the temperature fluctuation sequence includes: calculating the longitudinal vertical distance between any peak point in the temperature change curve and the nearest valley point in the future time to obtain an amplitude characteristic value; the amplitude characteristic value reflects the temperature difference between the peak point and the nearest valley point in the future time. The larger the amplitude characteristic value, the more obvious the temperature change. All amplitude characteristic values ​​are sorted in chronological order to obtain a temperature fluctuation sequence.

[0029] Furthermore, when the amplitude characteristic value in the temperature fluctuation sequence shows an increasing trend, it means that the deposition phenomenon of the solid catalyst is gradually becoming more serious, and the subsequent temperature is easily affected by the deposition and aggregation of the catalyst; on the contrary, when the amplitude characteristic value in the temperature fluctuation sequence does not show an increasing trend, it means that the deposition phenomenon is gradually dissociating, and the subsequent temperature will not be affected by the deposition and aggregation of the catalyst. Therefore, the deposition change degree value is obtained according to the similar characteristics of the data changes of the temperature fluctuation sequence and the preset ascending sequence; preferably, in an embodiment of the present invention, the step of obtaining the deposition change degree value includes: calculating the Spearman correlation coefficient of the temperature fluctuation sequence and the preset ascending sequence and positively mapping them to obtain the deposition change degree value; in an embodiment of the present invention, the preset ascending sequence is When the temperature fluctuation sequence is more similar to the data change trend of the preset ascending sequence, the Spearman correlation coefficient is closer to 1. When the data change trend is more opposite, the Spearman correlation coefficient is closer to -1. It should be noted that the calculation of the Spearman correlation coefficient belongs to the prior art, and the specific steps are not repeated here. The range of the sedimentation change degree value after positive correlation mapping is ; The larger the sedimentation change value, the more serious the sedimentation aggregation phenomenon is, and the subsequent temperature is more easily affected by the sedimentation phenomenon.

[0030] Step S4, adaptively adjusting the original lag coefficient weight of the abnormal change point in the ARIMA prediction model according to the deposition change degree value and the reaction deposition characteristic value to obtain an adaptive lag coefficient weight; predicting the temperature at a future moment according to the ARIMA prediction model including the adaptive lag coefficient weight to obtain a predicted temperature; and monitoring the production process according to the predicted temperature.

[0031] After obtaining the reaction deposition characteristic value and the deposition change degree value, the original lag coefficient weight of the abnormal change point in the ARIMA prediction model can be adaptively adjusted according to the deposition change degree value and the reaction deposition characteristic value to obtain an adaptive lag coefficient weight; preferably, in an embodiment of the present invention, the step of obtaining the adaptive lag coefficient weight includes: when the reaction deposition characteristic value does not exceed the preset deposition threshold, the adaptive lag coefficient weight of the abnormal change point is a constant 0; it means that the reactor may be heated unstably, so the adaptive lag coefficient weight of the abnormal change point is set to 0, so that the abnormal change point does not participate in the ARIMA temperature prediction for the future, and the prediction result is close to the temperature characteristics when the normal temperature trend changes, and the actual temperature value at the future is affected by the unstable heating of the reactor, so that there is a large difference between the predicted temperature and the actual temperature, and the temperature abnormality in the condensation process is monitored.

[0032] Furthermore, when the deposition change value does not exceed the preset change threshold, it means that the deposition phenomenon is gradually dissociating, and the subsequent temperature is not affected by the deposition. Therefore, it is necessary to reduce the hysteresis coefficient weight of the deposition feature point to reduce the influence of the deposition feature point on the prediction result; in the embodiment of the present invention, the preset change threshold is 0.5, which can be determined by the implementer according to the implementation scenario. Then, the product of the original hysteresis coefficient weight of the abnormal data point and the deposition change value is calculated to obtain the adaptive hysteresis coefficient weight; the original linear coefficient weight of all data points in the historical temperature time series is 1, and the adaptive hysteresis coefficient weight of such deposition feature points is less than the constant 1, which reduces the influence on the prediction result and makes the predicted temperature value closer to the actual temperature value. When the deposition change value exceeds the preset change threshold, it means that the deposition phenomenon increases, and the temperature will continue to be affected by the deposition of the solid catalyst in the future, so it is necessary to increase the importance of the deposition feature point in the prediction process. Calculate the sum of the deposition change value and the preset second constant to obtain the adjustment amount; in the embodiment of the present invention, the preset second constant is 1, which can be determined by the implementer according to the implementation scenario, in order to make the adaptive hysteresis coefficient weight exceed the original value. The product of the adjustment amount and the original hysteresis coefficient weight is calculated to obtain the adaptive hysteresis coefficient weight; the adaptive hysteresis coefficient weight of such deposition feature points exceeds the original hysteresis coefficient weight, and thus the temperature characteristics of the deposition feature points are more considered in the prediction process; the prediction results are closer to the temperature level caused by the deposition situation, and thus the prediction results are closer to the actual values; the situation in which the temperature is considered abnormal due to the large deviation between the prediction results and the actual values ​​caused by the deposition phenomenon is avoided.

[0033] Furthermore, the temperature at a future moment can be predicted according to the ARIMA prediction model including the adaptive hysteresis coefficient weight to obtain the predicted temperature; it should be noted that the ARIMA prediction model belongs to the prior art, and the specific prediction steps are not repeated here. In the embodiment of the present invention, the temperature value 10 seconds after the current moment is predicted. Finally, the production process is monitored according to the predicted temperature; the absolute value of the difference between the predicted temperature and the actual measured temperature at the same moment is calculated to obtain the temperature error value; when the temperature error value exceeds the preset error threshold, the methyl methacrylate production process is abnormal; the implementer can set the preset error threshold according to the implementation scenario. In the embodiment of the present invention, the preset error threshold is 10 degrees Celsius.

[0034] In summary, the embodiment of the present invention provides a method for monitoring data of methyl methacrylate production environment; obtain abnormal change points according to the data difference characteristics of the historical temperature time series; obtain reaction deposition characteristic values ​​according to the temperature difference characteristics of adjacent abnormal change points; determine whether the abnormal change point is a deposition characteristic point according to the reaction deposition characteristic value; obtain a temperature fluctuation sequence according to all deposition characteristic points; obtain a deposition change degree value according to the similarity characteristics of the temperature fluctuation sequence and the preset ascending sequence. The present invention obtains the adaptive hysteresis coefficient weight of the abnormal change point according to the deposition change degree value and the reaction deposition characteristic value; predicts the temperature according to the ARIMA prediction model containing the adaptive hysteresis coefficient weight; monitors the production process according to the predicted temperature, avoids the influence of the solid catalyst deposition phenomenon on temperature monitoring, and improves the accuracy of temperature monitoring.

[0035] It should be noted that the sequence of the above embodiments of the present invention is only for description and does not represent the advantages and disadvantages of the embodiments. The processes depicted in the accompanying drawings do not necessarily require the specific order or continuous order shown to achieve the desired results. In some embodiments, multitasking and parallel processing are also possible or may be advantageous.

[0036] The various embodiments in this specification are described in a progressive manner, and the same or similar parts between the various embodiments can be referenced to each other, and each embodiment focuses on the differences from other embodiments.

Claims

1. A method for monitoring methyl methacrylate production environment data, characterized in that: The method comprises the following steps: Get the historical temperature time series of the reactor at the current moment in the production process of methyl methacrylate; Obtain an abnormal change point according to the data difference characteristics between the data point and the adjacent data points in the historical temperature time series, and the difference characteristics between the data point and the overall data change trend; obtain a reaction deposition characteristic value according to the temperature difference characteristics of the adjacent abnormal change points; According to the reaction deposition characteristic value, determine whether the abnormal change point is a deposition characteristic point; if there is a deposition characteristic point, construct a temperature change curve according to all deposition characteristic points, and obtain a temperature fluctuation sequence according to the temperature difference characteristics of the peaks and valleys in the temperature change curve; obtain a deposition change degree value according to the similarity characteristics of the temperature fluctuation sequence and the data change of the preset ascending sequence; The original lag coefficient weight of the abnormal change point in the ARIMA prediction model is adaptively adjusted according to the deposition change degree value and the reaction deposition characteristic value to obtain an adaptive lag coefficient weight; the temperature at a future moment is predicted according to the ARIMA prediction model including the adaptive lag coefficient weight to obtain a predicted temperature; and the production process is monitored according to the predicted temperature.

2. A method for monitoring methyl methacrylate production environment data according to claim 1, characterized in that: The step of obtaining abnormal change points according to the data difference characteristics between the data point and the adjacent data points in the historical temperature time series and the difference characteristics between the data point and the overall data change trend includes: Constructing a temperature-time coordinate system of the historical temperature time series, calculating the sum of the Euclidean distances between any data point in the historical temperature time series and the adjacent data points before and after in the temperature-time coordinate system, and obtaining a neighborhood difference characteristic value of the arbitrary data point; performing time series decomposition on the historical temperature time series in the temperature-time coordinate system, obtaining a temperature change trend curve, calculating and normalizing the Euclidean distance between the arbitrary data point and the data point at the same time in the temperature change trend curve, and obtaining a trend difference characteristic value of the arbitrary data point; calculating and normalizing the inverse of the neighborhood difference characteristic value, and obtaining a neighborhood similarity; calculating the average value of the trend difference characteristic value and the neighborhood similarity, and obtaining the degree of abnormality of the arbitrary data point; When the abnormal degree of the arbitrary data point exceeds a preset abnormal threshold, the arbitrary data point is an abnormal change point.

3. A method for monitoring methyl methacrylate production environment data according to claim 1, characterized in that: The step of obtaining the reaction deposition characteristic value according to the temperature difference characteristics of adjacent abnormal change points comprises: Calculate the temperature difference between any abnormal change point and the next adjacent abnormal change point to obtain the temperature difference; calculate the inverse of the absolute value of the sum of all temperature differences to obtain the temperature stability characteristic value; calculate the ratio of the number of abnormal change points to a preset constant to obtain the quantity characteristic value; calculate the product of the quantity characteristic value and the temperature stability characteristic value to obtain the reaction deposition characteristic value.

4. A method for monitoring methyl methacrylate production environment data according to claim 1, characterized in that: The step of judging whether the abnormal change point is a deposition feature point according to the reaction deposition feature value comprises: When the reaction deposition characteristic value exceeds a preset deposition threshold, the abnormal change point is a deposition characteristic point.

5. A method for monitoring methyl methacrylate production environment data according to claim 1, characterized in that: The step of obtaining a temperature fluctuation sequence according to the temperature difference characteristics of the peaks and valleys in the temperature change curve comprises: The longitudinal vertical distance between any peak point in the temperature change curve and the nearest trough point in the future time is calculated to obtain the amplitude characteristic value; all the amplitude characteristic values ​​are sorted in chronological order to obtain the temperature fluctuation sequence.

6. A method for monitoring methyl methacrylate production environment data according to claim 1, characterized in that: The step of obtaining the deposition variation value according to the similar characteristics of the data variation of the temperature fluctuation sequence and the preset ascending sequence comprises: The Spearman correlation coefficient between the temperature fluctuation sequence and the preset ascending sequence is calculated and positively correlated to obtain the deposition change degree value.

7. A method for monitoring methyl methacrylate production environment data according to claim 4, characterized in that: The step of adaptively adjusting the original hysteresis coefficient weight of the abnormal change point in the ARIMA prediction model according to the deposition change degree value and the reaction deposition characteristic value to obtain the adaptive hysteresis coefficient weight comprises: When the reaction deposition characteristic value does not exceed the preset deposition threshold, the adaptive hysteresis coefficient weight of the abnormal change point is a constant of 0; when the deposition change degree value does not exceed the preset change threshold, the product of the original hysteresis coefficient weight of the abnormal data point and the deposition change degree value is calculated to obtain the adaptive hysteresis coefficient weight; when the deposition change degree value exceeds the preset change threshold, the sum of the deposition change degree value and a preset second constant is calculated to obtain the adjustment amount; the product of the adjustment amount and the original hysteresis coefficient weight is calculated to obtain the adaptive hysteresis coefficient weight.

8. A method for monitoring methyl methacrylate production environment data according to claim 1, characterized in that: The step of monitoring the production process according to the predicted temperature comprises: The absolute value of the difference between the predicted temperature and the actual measured temperature at the same time is calculated to obtain a temperature error value; when the temperature error value exceeds a preset error threshold, an abnormality occurs in the methyl methacrylate production process.

Images