Method and system for monitoring running state of toothpaste production equipment

Abstract

The invention relates to the technical field of data processing, in particular to a toothpaste production equipment running state monitoring method and system. The method comprises the following steps: acquiring a weight correction factor of each piece of parameter data according to a deviation condition of an initial LOF value between each piece of parameter data and historical parameter data; according to the gradient change condition of the initial LOF value of each piece of historical parameter data, obtaining a correction attenuation factor of each piece of historical parameter data of each piece of parameter data; correcting each piece of parameter data according to the correction attenuation factor and the weight correction factor to obtain a corrected LOF value of each piece of parameter data; and monitoring the running state of the toothpaste production equipment based on the corrected LOF value of each parameter data in each dimension parameter data sequence. According to the invention, the accuracy of monitoring the running state of the toothpaste production equipment is improved.

Description

Technical Field

[0001] The present invention relates to the technical field of data processing, and particularly to a method and system for monitoring the running state of toothpaste production equipment. Background Art

[0002] Toothpaste, as a daily consumer product, its production process involves multi-link collaboration (such as paste making, filling, and packaging). The stability of the equipment directly affects product quality and production efficiency. With the development of technology, toothpaste production equipment has achieved partial automation and accumulated a large amount of operation data, but lacks intelligent analysis means. For example, historical data such as motor current and vacuum degree recorded by the PLC control system is difficult to discover abnormal states during equipment operation and convert them into fault warning bases if potential associations are not mined through algorithms.

[0003] The Local Outlier Factor (LOF) algorithm can identify abnormal states during equipment operation by comparing the density differences between data points and their neighborhoods. For example, in toothpaste filling equipment, if the local density of parameters such as temperature, pressure, and flow rate significantly deviates from the normal range, the LOF algorithm can quickly locate potential fault points and avoid shutdowns caused by equipment overheating or wear.

[0004] However, toothpaste production involves the collaborative work of parameter data in multiple dimensions such as temperature, pressure, and flow rate, and there is a non-linear coupling relationship between parameter data (for example, an increase in temperature may change the viscosity of the material, which in turn affects the flow rate). The LOF value is calculated based on the parameter space distribution. Also, due to the dynamic drift of the original parameters of toothpaste production equipment caused by equipment aging, raw material batch differences, or environmental fluctuations (such as workshop temperature), the LOF value of equipment parameter data deviates from the true abnormal state, and directly using the uncorrected LOF value will reduce the accuracy of monitoring the running state of toothpaste production equipment. Summary of the Invention

[0005] In order to solve the technical problem that due to the collaborative work of parameter data in multiple dimensions such as temperature, pressure, and flow rate in toothpaste production, there is a non-linear coupling relationship between parameter data, resulting in the LOF value of equipment parameter data deviating from the true abnormal state, and directly using the uncorrected LOF value will reduce the accuracy of monitoring the running state of toothpaste production equipment, the present invention provides a method and system for monitoring the running state of toothpaste production equipment.

[0006] In the first aspect, the present invention provides a method for monitoring the running state of toothpaste production equipment, adopting the following technical solution: A method for monitoring the running state of toothpaste production equipment includes the steps of: Collecting multi-dimensional parameter data sequences during the operation of toothpaste production equipment; Obtain the initial LOF value of each parameter data in each dimension parameter data sequence; obtain all historical parameter data of each parameter data; obtain the weight correction factor of each parameter data according to the deviation of the initial LOF value between each parameter data and the historical parameter data; obtain the correction attenuation factor of each historical parameter data of each parameter data according to the gradient change of the initial LOF value of each historical parameter data; Correct each parameter data according to the correction attenuation factor and the weight correction factor to obtain the corrected LOF value of each parameter data; Monitor the running state of the toothpaste production equipment based on the corrected LOF value of each parameter data in each dimension parameter data sequence.

[0007] The innovation of the present invention is to obtain the weight correction factor of each parameter data according to the deviation of the initial LOF value between each parameter data and the historical parameter data; dynamically adjust the correction weight factor accordingly to make the correction result more accurate; obtain the correction attenuation factor of each historical parameter data of each parameter data according to the gradient change of the initial LOF value of each historical parameter data; this can reduce their excessive influence on the current correction and ensure that more reliance is placed on the latest and more relevant parameter data during monitoring; correct each parameter data according to the correction attenuation factor and the weight correction factor to obtain the corrected LOF value of each parameter data; the corrected LOF value can more accurately reflect the local outlier situation of the current parameter data relative to the historical parameter data; monitor the running state of the toothpaste production equipment based on the corrected LOF value, thereby improving the accuracy of monitoring the running state of the toothpaste production equipment.

[0008] Preferably, the obtaining all historical parameter data of each parameter data includes: Preset a historical parameter m, and in any dimension parameter data sequence during the operation of the toothpaste production equipment, record the m parameter data before the t-th parameter data as the historical parameter data of the t-th parameter data.

[0009] Preferably, the obtaining the weight correction factor of each parameter data according to the deviation of the initial LOF value between each parameter data and the historical parameter data includes: Obtain the offset value of each parameter data; Take the inverse normalization value of the ratio between the offset value of the t-th parameter data and the standard deviation of the initial LOF values of all historical parameter data of the t-th parameter data as the weight correction factor of the t-th parameter data.

[0010] Beneficial effects: By calculating the deviation of the initial LOF value between the current parameter data and the historical parameter data, the difference in the initial LOF value between the current parameter data and the historical parameter data can be effectively identified, thereby dynamically adjusting the correction weight factor to make the correction result more accurate and avoiding over-reliance on the initial LOF value of the historical parameter data that does not conform to the current trend.

[0011] Preferably, obtaining the offset value of each parameter data includes: Taking the absolute value of the difference between the mean of the initial LOF values of all historical parameter data of the t-th parameter data and the initial LOF value of the t-th parameter data as the offset value of the t-th parameter data.

[0012] Preferably, obtaining the correction attenuation factor of each historical parameter data of each parameter data according to the gradient change of the initial LOF value of each historical parameter data includes: Obtaining the gradient value of the initial LOF value of each historical parameter data of each parameter data; For the -th historical parameter data, taking the normalized value of the product of the gradient of the -th historical parameter data and the sequence number value of the -th historical parameter data as the correction attenuation factor of the

[0013] Beneficial effects: By attenuating the weights of the historical parameter data of each parameter data, their excessive influence on the current correction can be reduced, ensuring more reliance on the latest and more relevant parameter data during monitoring.

[0014] Preferably, obtaining the gradient value of the initial LOF value of each historical parameter data of each parameter data includes: For the b-th historical parameter data of the t-th parameter data, among all the historical parameter data of the b-th historical parameter data, taking the absolute value of the difference between the b-th historical parameter data and the (b + 1)-th historical parameter data as the first difference of the b-th historical parameter data; taking the cumulative sum of the first differences of all the historical parameter data of the -th historical parameter data as the gradient value of the initial LOF value of the -th historical parameter data. -th historical parameter data; -th historical parameter data; -th historical parameter data.

[0015] Preferably, correcting each parameter data according to the correction attenuation factor and the weight correction factor to obtain the corrected LOF value of each parameter data includes: Obtaining the first correction value according to the weight correction factor; Obtaining the second correction value according to the correction attenuation factor; Use the sum of the first correction value and the second correction value as the corrected LOF value of the t-th parameter data.

[0016] Beneficial effects: The corrected LOF value can more accurately reflect the local outlier situation of the current parameter data relative to the historical parameter data; if the influence of the historical parameter data is too large, the correction attenuation factor can effectively weaken its influence, ensuring that the corrected LOF value can more accurately capture abnormal data.

[0017] Preferably, obtaining the first correction value according to the weight correction factor includes: Denote the product of the weight correction factor of the t-th parameter data and the initial LOF value of the t-th parameter data as the first correction value.

[0018] Preferably, obtaining the second correction value according to the correction attenuation factor includes: Denote the difference between 1 and the weight correction factor of the t-th parameter data as the correction difference; denote the product of the correction attenuation factor of the th historical parameter data of the t-th parameter data and the initial LOF value of the th historical parameter data as the first product of the th historical parameter data; denote the product of the sum of the first products of all historical parameter data of the t-th parameter data and the correction difference as the second correction value.

[0019] Beneficial effects: As the parameter data is continuously updated, the historical parameter data may cause unnecessary interference to the judgment of the monitoring system. By means of the correction attenuation factor, the influence of these outdated historical parameter data can be effectively reduced, making the monitoring system more dependent on the parameter data that is relevant to the current parameter data.

[0020] In a second aspect, the present invention provides a monitoring system for the operating state of a toothpaste production device, adopting the following technical solution: A monitoring system for the operating state of a toothpaste production device includes: a processor and a memory. The memory stores computer program instructions, and when the computer program instructions are executed by the processor, the above-mentioned method for monitoring the operating state of a toothpaste production device is implemented.

[0021] By adopting the above technical solution, the above-mentioned method for monitoring the operating state of a toothpaste production device is generated into a computer program and stored in the memory to be loaded and executed by the processor, so as to manufacture a terminal device according to the memory and the processor, which is convenient to use.

[0022] The present invention has the following technical effects: According to the deviation of each parameter data from the initial LOF value of the historical parameter data, the weight correction factor of each parameter data is obtained; the correction weight factor is dynamically adjusted in this way to make the correction result more accurate; according to the gradient change of the initial LOF value of each historical parameter data, the correction attenuation factor of each historical parameter data for each parameter data is obtained; in this way, their excessive influence on the current correction can be reduced, ensuring that more reliance is placed on the latest and more relevant parameter data during monitoring; according to the correction attenuation factor and the weight correction factor, each parameter data is corrected to obtain the corrected LOF value of each parameter data; the corrected LOF value can more accurately reflect the local outlier situation of the current parameter data relative to the historical parameter data; the operating state of the toothpaste production equipment is monitored based on the corrected LOF value, thereby improving the accuracy of monitoring the operating state of the toothpaste production equipment. BRIEF DESCRIPTION OF THE DRAWINGS

[0023] Figure 1 is a flowchart of a method for monitoring the operating state of a toothpaste production equipment according to an embodiment of the present invention. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0024] Next, the technical solutions in the embodiments of the present invention will be clearly and completely described in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are part of the embodiments of the present invention, rather than all of the embodiments.

[0025] An embodiment of the present invention discloses a method for monitoring the operating state of a toothpaste production equipment. Referring to Figure 1 , it includes steps S1 - S4: S1: Collect a sequence of parameter data in multiple dimensions during the operation of the toothpaste production equipment.

[0026] In the implementation of the present invention, the preset sampling frequency is 1 second / time. Temperature sensors, pressure sensors, and flow sensors are respectively installed in the heating pipeline, the inlet and outlet of the storage tank, and the raw material pipeline of the toothpaste production equipment. During the operation of the toothpaste production equipment, temperature data, pressure data, and flow data are collected at each moment for a total of two hours; The temperature data, pressure data, and flow data of each sampling moment are used as a sequence of parameter data in multiple dimensions during the operation of the toothpaste production equipment.

[0027] S2: Obtain the initial LOF value of each parameter data in each dimension parameter data sequence; obtain all historical parameter data of each parameter data; obtain the weight correction factor of each parameter data according to the deviation of the initial LOF value between each parameter data and the historical parameter data; obtain the correction attenuation factor of each historical parameter data of each parameter data according to the gradient change of the initial LOF value of each historical parameter data.

[0028] It should be noted that toothpaste production involves the collaborative work of various dimension parameter data such as temperature, pressure, and flow rate, and there is a non-linear coupling relationship between the parameter data (for example, an increase in temperature may change the viscosity of the material, thereby affecting the flow rate); the LOF value is calculated based on the parameter space distribution, and because the original parameters of the toothpaste production equipment may have dynamic drift due to equipment aging, raw material batch differences, or environmental fluctuations (such as workshop temperature), resulting in the deviation of the LOF value of the equipment parameter data from the true abnormal state, directly using the uncorrected LOF value will reduce the accuracy of the operation status monitoring of the toothpaste production equipment; therefore, a correction mechanism needs to be introduced to correct the LOF values of various dimension parameter data during the operation of the toothpaste production equipment to ensure its accuracy and avoid misjudgment.

[0029] The embodiment of the present invention is described by taking any one of the dimension parameter data during the operation of the toothpaste production equipment as an example; In the embodiment of the present invention, the LOF algorithm is used to obtain the LOF value of each parameter data in the above-mentioned any one of the dimension parameter data sequences during the operation of the toothpaste production equipment, and it is recorded as the initial LOF value of each parameter data.

[0030] Among them, the LOF algorithm is a prior art, and no more details are given here in this embodiment.

[0031] It should be noted that the LOF value is used to measure the comparison of the density of a data point in a local area with the density of its neighborhood; if the LOF value of a certain data point deviates significantly from the LOF value of the historical data, it may indicate that there is an abnormality at this point; when an abnormal LOF value of a certain data point is found, it needs to be corrected based on the historical data to ensure that it is within a suitable range; in the corresponding correction method, the weight of the data point should be appropriately reduced to avoid excessive influence of abnormal fluctuations on the overall data analysis during the subsequent correction process; by reducing the weight, the influence of the abnormal point can be reduced, and the cumulative amplification effect on the overall data can be avoided.

[0032] In the embodiment of the present invention, the specific method for obtaining the weight correction factor of each parameter data according to the deviation of the initial LOF value between each parameter data and the historical parameter data is as follows: Preset a historical parameter m = 5. In any sequence of dimensional parameter data during the operation of the toothpaste production equipment, the m parameter data before the t-th parameter data are all recorded as the historical parameter data of the t-th parameter data; The absolute value of the difference between the mean of the initial LOF values of all historical parameter data of the t-th parameter data and the initial LOF value of the t-th parameter data is denoted as the offset value of the t-th parameter data; the inverse proportional normalization value of the ratio of the offset value of the t-th parameter data to the standard deviation of the initial LOF values of all historical parameter data of the t-th parameter data is used as the weight correction factor of the t-th parameter data; The specific formula is:

[0033] In the formula, represents the weight correction factor of the t-th parameter data; represents the initial LOF value of the t-th parameter data; represents the mean of the initial LOF values of all historical parameter data of the t-th parameter data; represents a preset hyperparameter, which is preset in this implementation , used to prevent the denominator from being zero; represents the exponential function with the natural constant as the base. The embodiment uses model to present the inverse proportional relationship and normalization process, is the input of the model, and the implementer can select the inverse proportional function and normalization function according to the actual situation.

[0034] Among them, represents the degree of deviation of the t-th parameter data; when the initial LOF value of the t-th parameter data deviates greatly from the mean of the initial LOF values of all its historical parameter data, it indicates that there is a certain abnormal data fluctuation in the t-th parameter data, that is, the possibility of distortion of the initial LOF value of the t-th parameter data is relatively large. Then, the initial LOF value of the t-th parameter data needs to be corrected to a large extent by the initial LOF values of the historical parameter data of the t-th parameter data, but the weight parameter corresponding to the initial LOF value of the t-th parameter data needs to be reduced to avoid this deviation being accumulated and amplified in the subsequent correction process.

[0035] It should be noted that when the initial LOF value gradient of all historical parameter data of a certain parameter data is relatively large, it means that its historical parameter data shows an upward trend, which may indicate that the parameter data of this dimension of the past toothpaste production equipment has changed greatly or there are abnormal fluctuations; in this case, the current correction work should rely more on recent data because recent data can better reflect the current operating state of the equipment, so it is necessary to accelerate the attenuation of historical data; the attenuation factor determines the influence degree of historical data on the correction of current data, which means the higher the attention to recent data.

[0036] In the embodiment of the present invention, according to the gradient change of the initial LOF value of each historical parameter data, the specific method for obtaining the correction attenuation factor of each historical parameter data of each parameter data is as follows: For the th historical parameter data of the t-th parameter data, among all the historical parameter data of the th historical parameter data, the absolute value of the difference between the th historical parameter data of the b-th historical parameter data and the (b + 1)-th historical parameter data is denoted as the first difference of the b-th historical parameter data; the accumulated sum of the first differences of all the historical parameter data of the th historical parameter data is used as the gradient value of the initial LOF value of the th historical parameter data; The normalized value of the product of the gradient of the th historical parameter data and the sequence number value of the th historical parameter data is used as the correction attenuation factor of the th historical parameter data; The specific formula is:

[0037] In the formula, represents the correction attenuation factor of the th historical parameter data of the t-th parameter data; represents the sequence number of the th historical parameter data of the t-th parameter data; represents the gradient value of the initial LOF value of the th historical parameter data of the t-th parameter data; represents the linear normalization function.

[0038] S3: According to the correction attenuation factor and the weight correction factor, correct each parameter data to obtain the corrected LOF value of each parameter data.

[0039] It should be noted that the operating parameters of the toothpaste production equipment (such as temperature, pressure, and flow rate) may fluctuate due to factors such as raw material batches, environmental temperature and humidity, and equipment wear. The exponential smoothing method can quickly capture the short-term change trend of parameter data by assigning higher weights to recent data and lower weights to distant data, avoiding the lag in LOF value correction caused by interference from historical parameter data. For example, when the flow rate parameter is abnormal due to a change in the viscosity of the raw material in the equipment, the exponential smoothing can quickly adjust the weights to make the corrected LOF value closer to the current state.

[0040] In the embodiments of the present invention, the specific method for correcting each parameter data according to the correction attenuation factor and the weight correction factor to obtain the corrected LOF value of each parameter data is as follows: Denote the product of the weight correction factor of the t-th parameter data and the initial LOF value of the t-th parameter data as the first correction value; Denote the difference between 1 and the weight correction factor of the t-th parameter data as the correction difference; denote the product of the correction attenuation factor of the r-th historical parameter data of the t-th parameter data and the initial LOF value of the r-th historical parameter data as the first product of the r-th historical parameter data; denote the product of the sum of the first products of all historical parameter data of the t-th parameter data and the correction difference as the second correction value; Take the sum of the first correction value and the second correction value as the corrected LOF value of the t-th parameter data; The specific formula is:

[0041] In the formula, represents the corrected LOF value of the t-th parameter data; represents the weight correction factor of the t-th parameter data; represents the initial LOF value of the t-th parameter data; represents the number of all historical parameter data of the t-th parameter data; represents the -th historical parameter data of the t-th parameter data; represents the -th historical parameter data of the t-th parameter data;

[0042] It should be noted that the above method corrects the initial LOF value of the t-th parameter data through the weight correction factor of the t-th parameter data and the correction attenuation factor of each of its historical parameter data; among them, the weight correction factor indicates the weight of the initial LOF value of the t-th parameter data in its corrected LOF value. The larger the weight correction factor, the higher the credibility of the LOF value of the t-th parameter data, and the larger the proportion of its initial LOF value in the correction process of the LOF value of the t-th parameter data, while the proportion of the initial LOF value of the historical parameter data is smaller; in addition, in in, control the overall weight of the initial LOF values of the historical parameter data, and the correction attenuation factor determines its internal distribution. The larger the correction attenuation factor of the historical parameter data, the more it indicates that the influence of the recent historical production parameters is emphasized during the correction process.

[0043] S4: Monitor the running state of the toothpaste production equipment based on the corrected LOF values of each parameter data in each dimension parameter data sequence.

[0044] In the embodiments of the present invention, the specific method for monitoring the running state of the toothpaste production equipment based on the corrected LOF values of each parameter data in each dimension parameter data sequence is as follows; Preset a threshold parameter T = 1. In other embodiments, the implementer can preset the value of the threshold parameter T according to the specific implementation situation; If the corrected LOF value of any parameter data in any dimension parameter data sequence is greater than the threshold parameter T, it indicates that there is an abnormal fluctuation in the data in the parameter data sequence, and there may be potential fault risks during the operation of the toothpaste production equipment, and relevant technical personnel need to be notified in time for repair and maintenance.

[0045] The above are all preferred embodiments of the present invention. Without restricting the protection scope of the present invention accordingly, therefore: All equivalent changes made according to the structure, shape, and principle of the present invention should be covered within the protection scope of the present invention.

Claims

1. A method for monitoring the operating state of a toothpaste production device, characterized in that, Including the steps: Collect a sequence of multi-dimensional parameter data during the operation of the toothpaste production equipment; Obtain the initial LOF value of each parameter data in each sequence of multi-dimensional parameter data; obtain all historical parameter data of each parameter data; according to the deviation of the initial LOF value between each parameter data and the historical parameter data, obtain the weight correction factor of each parameter data; According to the gradient change of the initial LOF value of each historical parameter data, obtain the correction attenuation factor of each historical parameter data of each parameter data; According to the correction attenuation factor and the weight correction factor, correct each parameter data to obtain the corrected LOF value of each parameter data; Monitor the running state of the toothpaste production equipment based on the corrected LOF value of each parameter data in each sequence of multi-dimensional parameter data.

2. A method for monitoring the operating state of a toothpaste production device according to claim 1, characterized in that, The obtaining all historical parameter data of each parameter data includes: Preset a historical parameter m. In any sequence of multi-dimensional parameter data during the operation of the toothpaste production equipment, the m parameter data before the t-th parameter data are all recorded as the historical parameter data of the t-th parameter data.

3. A method for monitoring the operating state of a toothpaste production device according to claim 1, characterized in that, The obtaining the weight correction factor of each parameter data according to the deviation of the initial LOF value between each parameter data and the historical parameter data includes: Obtain the offset value of each parameter data; Take the inverse proportional normalization value of the ratio between the offset value of the t-th parameter data and the standard deviation of the initial LOF values of all historical parameter data of the t-th parameter data as the weight correction factor of the t-th parameter data.

4. The method for monitoring the running state of a toothpaste production device according to claim 3, wherein, The obtaining the offset value of each parameter data includes: Record the absolute value of the difference between the mean of the initial LOF values of all historical parameter data of the t-th parameter data and the initial LOF value of the t-th parameter data as the offset value of the t-th parameter data.

5. A method for monitoring the operating state of a toothpaste production device according to claim 1, characterized in that, The obtaining the correction attenuation factor of each historical parameter data of each parameter data according to the gradient change of the initial LOF value of each historical parameter data includes: Obtain the gradient value of the initial LOF value of each historical parameter data of each parameter data; Normalize the product of the gradient of the th historical parameter data and the sequence number value of the th historical parameter data, and use it as the th historical parameter data's correction attenuation factor.

6. The method for monitoring the running state of a toothpaste production device according to claim 5, characterized in that, The obtaining the gradient value of the initial LOF value of each historical parameter data of each parameter data includes: For the th historical parameter data of the t-th parameter data, among all the historical parameter data of the th historical parameter data, the absolute value of the difference between the th historical parameter data's b-th historical parameter data and the b+1-th historical parameter data is denoted as the first difference of the b-th historical parameter data; the cumulative sum of the first differences of all the historical parameter data of the th historical parameter data is used as the gradient value of the initial LOF value of the th historical parameter data.

7. A method for monitoring the operating state of a toothpaste production device according to claim 1, characterized in that, The correcting each parameter data according to the correction attenuation factor and the weight correction factor to obtain the corrected LOF value of each parameter data includes: Obtain the first correction value according to the weight correction factor; Obtain the second correction value according to the correction attenuation factor; Take the sum of the first correction value and the second correction value as the corrected LOF value of the t-th parameter data.

8. A method for monitoring the operating state of a toothpaste production device according to claim 7, characterized in that, The obtaining the first correction value according to the weight correction factor includes: Record the product of the weight correction factor of the t-th parameter data and the initial LOF value of the t-th parameter data as the first correction value.

9. A method for monitoring the operating state of a toothpaste production device according to claim 7, characterized in that, The obtaining the second correction value according to the correction attenuation factor includes: Denote the difference between 1 and the weight correction factor of the t-th parameter data as the correction difference; denote the product of the correction attenuation factor of the -th historical parameter data of the t-th parameter data and the initial LOF value of the -th historical parameter data as the first product of the -th historical parameter data; denote the product of the sum of the first products of all historical parameter data of the t-th parameter data and the correction difference as the second correction value.

10. A monitoring system for the operating state of a toothpaste production device, characterized in that, Including: A processor and a memory, the memory stores computer program instructions, and when the computer program instructions are executed by the processor, a method for monitoring the running state of a toothpaste production equipment according to any one of claims 1-9 is implemented.

Images