Enzyme preparation production data analysis method and system based on big data

Abstract

The invention relates to the technical field of data processing, and particularly discloses an enzyme preparation production data analysis method and system based on big data, and the method comprises the following steps: obtaining a current temperature sequence, a historical abnormal temperature sequence and a historical normal temperature sequence of each temperature measurement point in all height regions of a fermentation tank; and calculating an abnormal score of enzyme preparation production, when the abnormal score is greater than a threshold value, determining that enzyme preparation production is abnormal, and controlling the fermentation tank to stop running. The enzyme preparation production data analysis method and system based on big data provided by the invention are more accurate in enzyme preparation production anomaly detection.

Description

Technical Field

[0001] The present invention relates to the technical field of data processing, and in particular to a method and system for analyzing enzyme preparation production data based on big data. Background Art

[0002] Since the method of producing enzyme preparations by microbial fermentation has the advantages of high efficiency, easy control, diverse products and high product quality, the use of this method is relatively common. In the process of producing enzyme preparations by microbial fermentation, it is necessary to inject culture fluid into the fermentation tank, inoculate the microorganisms producing enzymes, and then ferment under certain conditions to obtain the required enzyme preparations. It is particularly important to control the temperature during the fermentation process, because the temperature can affect the fermentation speed of the microorganisms and the activity of the enzyme preparation, and then affect the yield and quality of the enzyme preparation. Therefore, it is necessary to monitor the production temperature in real time to prevent the production temperature from being abnormal and affecting the production of the enzyme preparation. In the prior art, monitoring the production temperature is to set a large number of temperature measuring points at different positions of the fermentation tank, measure the temperature data at different temperature measuring points, and these temperature data have the characteristics of large scale and high processing speed requirements to realize real-time monitoring. Specifically, when the mean value of the temperature at each temperature measuring point is in the preset temperature range, it is determined that the production temperature is normal, that is, the enzyme preparation production process is normal, otherwise, it is determined that the production temperature is abnormal, that is, the enzyme preparation production process is abnormal.

[0003] However, in the production process of enzyme preparations, the metabolism of microorganisms will generate heat, and the oxygen content of different height areas of the fermentation tank is different, the metabolic rate of the microorganism is also different, and the heat generated by metabolism is also different, so that the temperature change of each height area is also different. Therefore, the temperature of different height areas has different influences on whether the production temperature is abnormal. In the prior art, the temperature of different height areas all adopts the same weight, which will make some abnormal temperatures and normal temperatures averaged, and the average temperature obtained is within the normal temperature range. However, the average temperature calculated by this method cannot accurately and truly reflect the production temperature, which will affect the production process of the enzyme preparation, reduce the yield and quality of the enzyme preparation, and affect the economic benefits. Summary of the invention

[0004] The present invention provides an enzyme preparation production data analysis method and system based on big data, aiming to solve the technical problem in the prior art that the weights of the temperatures in various height zones of a fermentation tank are the same, resulting in inaccurate abnormal detection results and affecting the production efficiency and quality of the enzyme preparation.

[0005] A method for analyzing enzyme preparation production data based on big data of the present invention comprises the following steps: Obtain the current temperature sequence, historical abnormal temperature sequence and historical normal temperature sequence of each temperature measuring point in all height areas of the fermentation tank; Among them, the historical abnormal temperature sequence and the historical normal temperature sequence are the temperatures at each moment in the same production stage as the current temperature sequence; Calculate the abnormal score of enzyme preparation production. When the abnormal score is greater than the threshold, it is determined that there is an abnormality in enzyme preparation production, and the fermentation tank is controlled to stop running; Among them, the anomaly score is the ratio of the anomaly mean of all height areas to the total number of height areas; the anomaly mean is obtained by weighting the corresponding anomaly degree using the weight of each height area; the anomaly degree is positively correlated with the similarity between the current temperature sequence of each temperature measuring point and the historical abnormal temperature sequence, and is negatively correlated with the similarity between the current temperature sequence of each temperature measuring point and the historical normal temperature sequence; the weight is the ratio of the temperature significance of the corresponding height area to the sum of the temperature significance of all height areas; the temperature significance is positively correlated with the difference between the temperature at the corresponding moment in the historical abnormal temperature sequence and the historical normal temperature sequence of the temperature measuring point, and is negatively correlated with the standard deviation of the difference between the temperature at all moments; the difference degree is the ratio of the absolute value of the difference between the temperature at the corresponding moment in the historical abnormal temperature sequence and the historical normal temperature sequence to the temperature at the corresponding moment in the historical normal temperature sequence.

[0006] In the above scheme, the anomaly score is obtained by weighting the corresponding anomaly degree by the weight of each height area. Since the influence of the temperature of different height areas on anomaly detection is taken into account, the calculation result of the anomaly score is more accurate, and the anomaly detection result is also more accurate.

[0007] Preferably, The degree of abnormality in the height area for: ; In the formula, For the In the height area The similarity between the current temperature series and the historical abnormal temperature series of each temperature measuring point, For the In the height area The similarity between the current temperature series of each temperature measuring point and the historical normal temperature series, For the The total number of temperature measurement points in the height area, is the standard normalization function.

[0008] In the above scheme, by comparing the similarity between the current temperature sequence and the historical abnormal temperature sequence with the similarity between the current temperature sequence and the historical normal temperature sequence, the abnormal degree of the altitude area can be intuitively reflected.

[0009] Preferably, The temperature significance of each altitude area for: ; In the formula, For the In the height area The historical abnormal temperature series and the historical normal temperature series of the temperature measuring points The temperature difference at each moment, For the In the height area The standard deviation of the difference between the temperature at all moments in the historical abnormal temperature series and the historical normal temperature series at each temperature measurement point, For the The total number of temperature measurement points in the height area, For the The length of the historical abnormal temperature sequence and the historical normal temperature sequence of the temperature measuring point in the height area, is the preset parameter, The value range is [0.01,0.02], is the standard normalization function.

[0010] In the above scheme, by comparing the temperature difference between the historical abnormal temperature sequence and the historical normal temperature sequence at corresponding moments and the standard deviation of the temperature difference at each moment, the reliability of the calculation result is higher, avoiding accidental errors that lead to errors in the calculation result.

[0011] Preferably, the calculation of the similarity comprises the following steps: The historical abnormal temperature sequence or the historical normal temperature sequence is set as the target temperature sequence, and the sequence of the target temperature sequence corresponding to the current temperature sequence and intercepted with the same length is taken as the target temperature subsequence; Get all cycles of the current temperature sequence and the target temperature subsequence; Take each period as the sliding window length and slide on the current temperature sequence and the target temperature subsequence with a step size of 1 to obtain the current window sequence and the target window sequence; The similarity between the current temperature sequence and the target temperature sequence at any temperature measurement point is inversely correlated with the absolute value of the difference between the mean of each current window sequence and the corresponding target window sequence.

[0012] In the above scheme, since the current window sequence and the target window sequence are compared at a smaller length, it is easier to capture similar subsequences in the current window sequence and the target window sequence, making the calculation result more accurate.

[0013] Preferably, the similarity between the current temperature sequence of any temperature measurement point and the target temperature sequence for: ; In the formula, The corresponding temperature measurement point is The period is the length of the sliding window. The mean of the current window sequence, The corresponding temperature measurement point is The period is the length of the sliding window. The mean of the target window sequence, is the total number of cycles of the current temperature sequence and the target temperature subsequence in the corresponding temperature measurement point, The corresponding temperature measurement point is The total number of current window sequences obtained by sliding window length is 1 period. The natural constant The exponential function of base .

[0014] In the above scheme, the natural constant The exponential function with as the base makes the value range of similarity be from 0 to 1, which is convenient for the subsequent calculation of the degree of abnormality.

[0015] Preferably, the similarity between the current temperature sequence of any temperature measurement point and the target temperature sequence for: ; In the formula, The corresponding temperature measurement point is The period is the length of the sliding window. The mean of the current window sequence, The corresponding temperature measurement point is The period is the length of the sliding window. The mean of the target window sequence, is the maximum value of the period of the current temperature sequence and the target temperature subsequence in the corresponding temperature measurement point, is the total number of cycles of the current temperature sequence and the target temperature subsequence in the corresponding temperature measurement point, The corresponding temperature measurement point is The total number of current window sequences obtained by sliding window length is 1 period. The natural constant The logarithmic function of the base.

[0016] Preferably, the periods of the current temperature sequence and the target temperature subsequence are obtained through Fourier series.

[0017] In the above scheme, the periods of the current temperature sequence and the target temperature subsequence are obtained by Fourier series, which is simple to calculate and easy to understand.

[0018] Preferably, the height region is obtained by equally dividing the fermenter along its height direction.

[0019] In the above scheme, the division of the height area of ​​the fermentation tank is relatively simple and convenient for calculation.

[0020] Preferably, the current temperature sequence is the temperature at each moment between the current moment and the start moment of the production stage in which it is located.

[0021] In the above scheme, by comparing the current temperature sequence with the historical abnormal temperature sequence and the historical normal temperature sequence in the same production stage as the current temperature sequence, the amount of calculation can be reduced and the calculation speed can be accelerated.

[0022] The present invention also provides an enzyme preparation production data analysis system based on big data, comprising a memory and a processor, wherein the processor executes a computer program stored in the memory to implement any of the enzyme preparation production data analysis methods based on big data described above.

[0023] The beneficial effects are: The scheme of the present invention obtains the weight of each height area of ​​the fermentation tank by the temperature significance of each height area of ​​the fermentation tank, and fully considers the influence of the temperature of different height areas on abnormal detection. In addition, the degree of abnormality of each height area is obtained by calculating the similarity between the current temperature sequence and the historical abnormal temperature sequence and the historical normal temperature sequence, and finally the weight of each height area is used to weight the corresponding abnormal degree to obtain the abnormal score, so that the calculation result is more accurate, and the abnormal detection of enzyme preparation production is more accurate. BRIEF DESCRIPTION OF THE DRAWINGS

[0024] Figure 1 A flowchart of the steps of the enzyme preparation production data analysis method based on big data according to an embodiment of the present invention; Figure 2 This is a structural block diagram of an enzyme preparation production data analysis system based on big data according to an embodiment of the present invention. DETAILED DESCRIPTION

[0025] The embodiments described below with reference to the accompanying drawings are exemplary and are intended to be used to explain the present invention, but should not be construed as limiting the present invention.

[0026] like Figure 1 As shown, according to a first aspect of the present invention, a method for analyzing enzyme preparation production data based on big data is provided, comprising the following steps: S1. Obtain the current temperature sequence, historical abnormal temperature sequence and historical normal temperature sequence of each temperature measuring point in all height areas of the fermentation tank.

[0027] Among them, the current temperature sequence is the temperature at each moment between the current moment and the start moment of the production stage in which it is located; the historical abnormal temperature sequence and the historical normal temperature sequence are the temperatures at each moment in the same production stage as the current temperature sequence.

[0028] The production of enzyme preparations is obtained by the metabolism of microorganisms in the fermentation tank, and microorganisms include various bacteria, fungi, etc. For example, Bacillus subtilis metabolism can produce protease and amylase, and Mucor metabolism can produce lipase, etc. The metabolic process of these microorganisms will produce heat, and the oxygen content of different height positions of the fermentation tank is different, the metabolic rate of the microorganism is also different, and the heat produced by metabolism is also different, so that the temperature change of each height region is also different. This causes the temperature of different height regions to have different influences on whether the production of enzyme preparations is abnormal. Therefore, it is necessary to collect the temperature of different height regions of the fermentation tank in step S1. The height region can be obtained by dividing the fermentation tank equally along its height direction. For example, the fermentation tank is evenly divided into three regions along the height direction, and a top region, a middle region and a bottom region can be obtained. Of course, the fermentation tank can also be equally divided into any number of height regions along the height direction.

[0029] A plurality of temperature measuring points are arranged in each height region, and the temperature measuring points can be evenly arranged along the circumference of the inner wall of the fermentation tank. A temperature sensor is arranged at each temperature measuring point to measure the temperature at the temperature measuring point.

[0030] Since the metabolic rates of microorganisms are different, the production of enzyme preparations also includes multiple production stages. Specifically, in the early stage of microbial fermentation, due to the sufficient culture fluid in the fermentation tank and the high oxygen content, the microorganisms accelerate metabolism and the enzyme preparation production rate is also fast; as the culture fluid in the fermentation tank decreases, the oxygen content decreases, the microbial metabolism tends to be stable, and the enzyme preparation production rate also tends to be stable; then as the culture fluid in the fermentation tank further decreases, the oxygen content further decreases, the microbial metabolism slows down, and the enzyme preparation production rate is also slow; finally, the microorganism stops metabolism and the enzyme preparation stops producing. Therefore, the production of enzyme preparations can include an accelerated production stage, a stable production stage, and a slow production stage. The degree of influence of temperature in each production stage on determining whether the enzyme preparation production is abnormal is also different. Therefore, it is necessary to obtain the historical abnormal temperature sequence and the historical normal temperature sequence in the same production stage as the current temperature sequence. The historical abnormal temperature sequence and the historical normal temperature sequence can be judged by the activity of the enzyme preparation. For example, when the activity of the enzyme preparation is greater than or equal to the preset activity, the production stage is judged to be normal, and the temperature sequence of the production stage is the historical normal temperature sequence; when the activity of the enzyme preparation is less than the preset activity, the production stage is judged to be abnormal, and the temperature sequence of the production stage is the historical abnormal temperature sequence.

[0031] In some embodiments, the acquired temperature sequence may be subjected to noise reduction processing to prevent noise interference from affecting the accuracy of subsequent calculation results. For example, a mean filter algorithm may be used to perform noise reduction processing on the temperature sequence.

[0032] S2. Calculate the abnormal score of enzyme preparation production. When the abnormal score is greater than a threshold, determine that there is an abnormality in the enzyme preparation production, and control the fermentation tank to stop running.

[0033] The anomaly score is the ratio of the anomaly mean of all height regions to the total number of height regions. The anomaly mean is obtained by weighting the corresponding anomaly degree using the weight of each height region.

[0034] This is because the greater the degree of abnormality in each height area, the more abnormal the enzyme production is, the greater the calculated abnormal mean is, and the higher the abnormal score is. In addition, different height areas have different weights for the degree of abnormality. The greater the weight, the higher the abnormal score. Therefore, it is necessary to calculate the degree of abnormality of each height area and its corresponding weight. In step S2, the acquisition of the abnormal score includes the following steps: S21. Calculate the degree of abnormality in each height area.

[0035] The degree of abnormality in any altitude area is positively correlated with the similarity between the current temperature sequence of each temperature measuring point and the historical abnormal temperature sequence, and negatively correlated with the similarity between the current temperature sequence of each temperature measuring point and the historical normal temperature sequence. This is because the more similar the current temperature sequence of the temperature measuring point and the historical abnormal temperature sequence are, the more likely the current temperature sequence is to cause abnormality in the altitude area, and the greater the degree of abnormality in the altitude area; conversely, the more similar the current temperature sequence of the temperature measuring point and the historical normal temperature sequence are, the less likely the current temperature sequence is to cause abnormality in the altitude area, and the smaller the degree of abnormality in the altitude area. In step S21, the calculation of the degree of abnormality also includes the following steps: S211, calculating the similarity between the current temperature sequence and the historical abnormal temperature sequence or the historical normal temperature sequence. Step S211 includes the following steps: S2111, setting a historical abnormal temperature sequence or a historical normal temperature sequence as a target temperature sequence, and taking a sequence of the target temperature sequence and the same length as the current temperature sequence as a target temperature subsequence; The purpose of obtaining the target temperature subsequence is that the current temperature sequence is the temperature at each moment between the current moment and the start moment of the production stage in which it is located. That is to say, the current temperature sequence is not a complete production stage, while the target temperature sequence is a complete production stage. Therefore, it is necessary to obtain the target temperature subsequence. It will be more accurate to calculate the similarity of two temperature sequences of equal length.

[0036] S2112. Obtain all cycles of the current temperature sequence and the target temperature subsequence.

[0037] During the production process of enzyme preparation, the agitator in the fermentation tank will stir, so that the temperature in the fermentation tank is relatively uniform, and the temperature fluctuation is small, which is beneficial to the production of enzyme preparation. Since the agitator has a constant rotation speed in each production stage, the current temperature sequence and the target temperature subsequence obtained have certain periodic changes. Therefore, the cycle of the current temperature sequence and the target temperature subsequence can be obtained by Fourier series. Specifically, each temperature in the current temperature sequence and the target temperature subsequence is plotted, and the waveform diagram of the current temperature sequence and the waveform diagram of the target temperature subsequence can be obtained, and each waveform diagram is formed by the superposition of multiple standard waves, such as a standard wave can be a sine wave or a cosine wave, so the waveform diagram can be decomposed into multiple standard waves using Fourier series, and then the cycle of each standard wave is obtained, as the cycle of the waveform diagram. Fourier series has the advantages of simple calculation and easy understanding.

[0038] S2113, slide on the current temperature sequence and the target temperature subsequence respectively with each period as the sliding window length and a step size of 1 to obtain a current window sequence and a target window sequence.

[0039] S2114: The similarity between the current temperature sequence and the target temperature sequence of any temperature measurement point is inversely correlated with the absolute value of the difference between the mean of each current window sequence and the corresponding target window sequence.

[0040] Then the similarity between the current temperature sequence of any temperature measurement point and the target temperature sequence is for: ; In the formula, The corresponding temperature measurement point is The period is the length of the sliding window. The mean of the current window sequence, The corresponding temperature measurement point is The period is the length of the sliding window. The mean of the target window sequence, is the total number of cycles of the current temperature sequence and the target temperature subsequence in the corresponding temperature measurement point, The corresponding temperature measurement point is The total number of current window sequences obtained by sliding window length is 1 period. The natural constant The exponential function of base .

[0041] In step S211, the similarity between the current temperature sequence and the target temperature sequence is calculated by comparing the difference between the mean values ​​of the current window sequence and the target window sequence in terms of period length. Since the current window sequence and the target window sequence are compared at a smaller length, it is easier to capture similar subsequences in the current window sequence and the target window sequence, making the calculation result more accurate.

[0042] In some alternative embodiments, the similarity between the current temperature sequence of any temperature measurement point and the target temperature sequence for: ; In the formula, The corresponding temperature measurement point is The period is the length of the sliding window. The mean of the current window sequence, The corresponding temperature measurement point is The period is the length of the sliding window. The mean of the target window sequence, is the maximum value of the period of the current temperature sequence and the target temperature subsequence in the corresponding temperature measurement point, is the total number of cycles of the current temperature sequence and the target temperature subsequence in the corresponding temperature measurement point, The corresponding temperature measurement point is The total number of current window sequences obtained by sliding window length is 1 period. The natural constant The logarithmic function of the base.

[0043] In some other alternative embodiments, the similarity between the current temperature sequence and the target temperature sequence may also be calculated using the Pearson correlation coefficient or the cosine similarity between the current temperature sequence and the target temperature subsequence.

[0044] S212. Obtain the abnormality degree of each height area.

[0045] The first The degree of abnormality in the height area for: ; In the formula, For the In the height area The similarity between the current temperature series and the historical abnormal temperature series of each temperature measuring point, For the In the height area The similarity between the current temperature series of each temperature measuring point and the historical normal temperature series, For the The total number of temperature measurement points in the height zone, is the standard normalization function. Among them, and Obtained from the similarity calculation formula in step S211.

[0046] In step S212, by comparing the similarity between the current temperature sequence and the historical abnormal temperature sequence with the similarity between the current temperature sequence and the historical normal temperature sequence, the abnormal degree of the altitude area can be intuitively reflected.

[0047] S22. Calculate the weight of the abnormality degree of each height area.

[0048] The weight of the abnormality of any altitude region is the ratio of the temperature significance of the altitude region to the sum of the temperature significance of all altitude regions. This is because the higher the temperature significance of each altitude region, the greater the influence of the temperature of the altitude region on the judgment of the abnormality of the altitude region, that is, the corresponding weight is also greater.

[0049] The temperature significance is positively correlated with the difference between the temperature at the corresponding moment in the historical abnormal temperature sequence and the historical normal temperature sequence of the temperature measuring point, and negatively correlated with the standard deviation of the temperature difference at all moments. This is because the greater the difference between the temperature at the corresponding moment in the historical abnormal temperature sequence and the historical normal temperature sequence, the greater the temperature significance of the height area due to the influence of temperature, which in turn leads to a greater corresponding weight for the height area. Therefore, step S22 also includes the following steps: S221. Calculate the temperature significance of any altitude area.

[0050] Step S221 includes the following steps: First, the difference between the temperature at the corresponding moment in the historical abnormal temperature sequence and the historical normal temperature sequence of any temperature measurement point is calculated. The difference is the ratio of the absolute value of the difference between the temperature at the corresponding moment in the historical abnormal temperature sequence and the historical normal temperature sequence to the temperature at the corresponding moment in the historical normal temperature sequence.

[0051] The degree of difference is: ; In the formula, For the In the height area The historical abnormal temperature series and the historical normal temperature series of the temperature measuring points The temperature difference at each moment, For the In the height area The number of historical abnormal temperature series in the temperature measuring point Temperature at a given moment, For the In the height area The first temperature in the historical normal temperature sequence of the temperature measuring point Temperature at a moment.

[0052] Then, calculate the temperature significance of any altitude area.

[0053] The first The temperature significance of each altitude area for: ; In the formula, For the In the height area The historical abnormal temperature series and the historical normal temperature series of the temperature measuring points The temperature difference at each moment, For the In the height area The standard deviation of the difference between the temperature at all moments in the historical abnormal temperature series and the historical normal temperature series at each temperature measurement point, For the The total number of temperature measurement points in the height area, For the The length of the historical abnormal temperature sequence and the historical normal temperature sequence of the temperature measuring point in the height area, is the preset parameter, The value range is [0.01,0.02], is the standard normalization function. The purpose is to prevent the denominator of the fraction from being 0 when the standard deviation is 0, resulting in inability to calculate.

[0054] In step S221, the temperature significance is characterized not only by comparing the temperature difference between the historical abnormal temperature sequence and the historical normal temperature sequence at corresponding moments, but also by the reciprocal of the standard deviation of the temperature difference at each moment to characterize the credibility. The smaller the standard deviation, the more consistent the temperature difference at each moment, the larger the reciprocal, the higher the credibility, thus avoiding accidental errors that lead to errors in the calculation results.

[0055] S222. Obtain the weight of the abnormality degree of each height area.

[0056] The first The weight of the abnormality degree of each height area for: ; In the formula, For the The temperature significance of each altitude area is is the total number of height zones.

[0057] In step S222, the greater the significance of the temperature in the height region, the greater the influence of the temperature in the height region on the anomaly score, and therefore a greater weight needs to be given to increase the attention paid to the height region.

[0058] S23. Obtain the abnormal score of enzyme preparation production.

[0059] The abnormal fraction of enzyme production for: ; In the formula, For the The degree of abnormality in the height area, For the The weight of the abnormality degree of each height area , is the total number of height zones.

[0060] After obtaining the abnormal score of enzyme preparation production, by comparing the abnormal score with the threshold, it can be determined whether there is an abnormality in enzyme preparation production. When the abnormal score is greater than the threshold, it is determined that there is an abnormality in enzyme preparation production, and the fermentation tank is controlled to stop running.

[0061] The threshold value range is [0.6, 0.8]. Preferably, the threshold value is 0.7. Of course, the threshold value can be adjusted as needed.

[0062] By stopping the fermentation tank, the enzyme preparation can be prevented from being inactivated due to abnormal conditions, which can cause losses. After the fermentation tank is stopped, maintenance personnel can be arranged to carry out maintenance.

[0063] In the enzyme preparation production data analysis method based on big data of the present invention, the degree of abnormality of each height area of ​​the fermentation tank is obtained by the similarity between the current temperature sequence and the historical abnormal temperature sequence and the similarity between the current temperature sequence and the historical normal temperature sequence, and the corresponding weight is obtained by the difference between the temperature at the corresponding moment of the historical abnormal temperature sequence and the historical normal temperature sequence, and then the abnormality score of the enzyme preparation production is obtained according to the degree of abnormality of each height area and the corresponding weight. Since the weights of different height areas are calculated to obtain the weights of different height areas in the calculation of the abnormal score, the calculation of the abnormal score is more accurate, and the weights of different height areas of the fermentation tank are avoided in the prior art because the weights of each height area of ​​the fermentation tank are the same, which leads to inaccurate calculation results, affects the production of enzyme preparations, reduces the yield and quality of enzyme preparations, and affects economic benefits.

[0064] like Figure 2As shown, according to the second aspect of the present invention, there is also provided an enzyme preparation production data analysis system based on big data, the system comprising a memory and a processor, the processor executing a computer program stored in the memory to implement the enzyme preparation production data analysis method based on big data described in the first aspect of the present invention.

[0065] The system also includes other components familiar to those skilled in the art, such as a communication bus and a communication interface, whose configuration and functions are known in the art and thus will not be described in detail here.

[0066] In the present invention, the aforementioned memory may be any tangible medium containing or storing a program, which may be used by or in combination with an instruction execution system, apparatus or device. For example, a computer-readable storage medium may be any appropriate magnetic storage medium or magneto-optical storage medium, such as a resistive random access memory RRAM (Resistive Random Access Memory), a dynamic random access memory DRAM (Dynamic Random Access Memory), a static random access memory SRAM (Static Random-Access Memory), an enhanced dynamic random access memory EDRAM (Enhanced Dynamic Random Access Memory), a high-bandwidth memory HBM (High-Bandwidth Memory), a hybrid memory cube HMC (Hybrid Memory Cube), etc., or any other medium that can be used to store the required information and can be accessed by an application, a module, or both. Any such computer storage medium may be part of a device or accessible or connectable to a device. Any application or module described in the present invention may be implemented by computer-readable / executable instructions stored or otherwise maintained by such a computer-readable medium.

[0067] Although this specification has shown and described a number of embodiments of the present invention, it will be apparent to those skilled in the art that such embodiments are provided by way of example only. Those skilled in the art will conceive of many modifications, changes and alternatives without departing from the ideas and spirit of the present invention. It should be understood that in the practice of the present invention, various alternatives to the embodiments of the present invention described herein may be employed.

Claims

1. A method for analyzing enzyme preparation production data based on big data, characterized in that: The method comprises the following steps: obtaining the current temperature sequence, the historical abnormal temperature sequence and the historical normal temperature sequence of each temperature measuring point in all height areas of the fermentation tank; wherein the historical abnormal temperature sequence and the historical normal temperature sequence are the temperatures at each moment in the same production stage as the current temperature sequence; Calculate the abnormal score of enzyme preparation production. When the abnormal score is greater than the threshold, it is determined that there is an abnormality in enzyme preparation production, and the fermentation tank is controlled to stop running; Among them, the anomaly score is the ratio of the anomaly mean of all height areas to the total number of height areas; the anomaly mean is obtained by weighting the corresponding anomaly degree using the weight of each height area; the anomaly degree is positively correlated with the similarity between the current temperature sequence of each temperature measuring point and the historical abnormal temperature sequence, and is negatively correlated with the similarity between the current temperature sequence of each temperature measuring point and the historical normal temperature sequence; the weight is the ratio of the temperature significance of the corresponding height area to the sum of the temperature significance of all height areas; the temperature significance is positively correlated with the difference between the temperature at the corresponding moment in the historical abnormal temperature sequence and the historical normal temperature sequence of the temperature measuring point, and is negatively correlated with the standard deviation of the difference between the temperature at all moments; the difference degree is the ratio of the absolute value of the difference between the temperature at the corresponding moment in the historical abnormal temperature sequence and the historical normal temperature sequence to the temperature at the corresponding moment in the historical normal temperature sequence.

2. The method for analyzing enzyme preparation production data based on big data according to claim 1, characterized in that: No. The degree of abnormality in the height area for: ； In the formula, For the In the height area The similarity between the current temperature series and the historical abnormal temperature series of each temperature measuring point, For the In the height area The similarity between the current temperature series of each temperature measuring point and the historical normal temperature series, For the The total number of temperature measurement points in the height area, is the standard normalization function.

3. The method for analyzing enzyme preparation production data based on big data according to claim 1, characterized in that: No. The temperature significance of each altitude area for: ； In the formula, For the In the height area The historical abnormal temperature series and the historical normal temperature series of the temperature measuring points The temperature difference at each moment, For the In the height area The standard deviation of the difference between the temperature at all moments in the historical abnormal temperature series and the historical normal temperature series at each temperature measurement point, For the The total number of temperature measurement points in the height area, For the The length of the historical abnormal temperature sequence and the historical normal temperature sequence of the temperature measuring point in the height area, is the preset parameter, The value range is [0.01,0.02], is the standard normalization function.

4. The method for analyzing enzyme preparation production data based on big data according to claim 1, characterized in that: The calculation of the similarity comprises the following steps: The historical abnormal temperature sequence or the historical normal temperature sequence is set as the target temperature sequence, and the sequence of the target temperature sequence corresponding to the current temperature sequence and intercepted with the same length is taken as the target temperature subsequence; Get all cycles of the current temperature sequence and the target temperature subsequence; Take each period as the sliding window length and slide on the current temperature sequence and the target temperature subsequence with a step size of 1 to obtain the current window sequence and the target window sequence; The similarity between the current temperature sequence and the target temperature sequence at any temperature measurement point is inversely correlated with the absolute value of the difference between the mean of each current window sequence and the corresponding target window sequence.

5. The method for analyzing enzyme preparation production data based on big data according to claim 4, characterized in that: The similarity between the current temperature sequence and the target temperature sequence at any temperature measurement point for: ； In the formula, The corresponding temperature measurement point is The period is the length of the sliding window. The mean of the current window sequence, The corresponding temperature measurement point is The period is the length of the sliding window. The mean of the target window sequence, is the total number of cycles of the current temperature sequence and the target temperature subsequence in the corresponding temperature measurement point, The corresponding temperature measurement point is The total number of current window sequences obtained by sliding window length is 1 period. The natural constant The exponential function of base .

6. The method for analyzing enzyme preparation production data based on big data according to claim 4, characterized in that: The similarity between the current temperature sequence and the target temperature sequence at any temperature measurement point for: ； In the formula, The corresponding temperature measurement point is The period is the length of the sliding window. The mean of the current window sequence, The corresponding temperature measurement point is The period is the length of the sliding window. The mean of the target window sequence, is the maximum value of the period of the current temperature sequence and the target temperature subsequence in the corresponding temperature measurement point, is the total number of cycles of the current temperature sequence and the target temperature subsequence in the corresponding temperature measurement point, The corresponding temperature measurement point is The total number of current window sequences obtained by sliding window length is 1 period. The natural constant The logarithmic function of the base.

7. The method for analyzing enzyme preparation production data based on big data according to claim 4, characterized in that: The periods of the current temperature sequence and the target temperature subsequence are obtained through Fourier series.

8. The method for analyzing enzyme preparation production data based on big data according to claim 1, characterized in that: The height region is obtained by equally dividing the fermentation tank along its height direction.

9. The method for analyzing enzyme preparation production data based on big data according to claim 1, characterized in that: The current temperature sequence is the temperature at each moment between the current moment and the start moment of the production stage in which it is located.

10. A big data-based enzyme preparation production data analysis system, comprising a memory and a processor, characterized in that: The processor executes the computer program stored in the memory to implement the enzyme preparation production data analysis method based on big data as described in any one of claims 1 to 9.

Images