A safety production management system for a semiconductor production line based on an internet of things

Abstract

The application relates to the technical field of temperature measurement, and discloses a semiconductor production line safety production management system based on the Internet of Things, which comprises the following steps: collecting temperature data of each position at several moments in multiple processes to obtain a simultaneous time temperature set composed of temperature data of different processes at the same moment of each position; obtaining a demarcation temperature of the simultaneous time temperature set at each moment of each position through inter-class variance; dividing an upper limit temperature set and a lower limit temperature set according to the demarcation temperature and obtaining a basic threshold proportion of the upper limit temperature set and the lower limit temperature set; obtaining a high-temperature degree at each moment of each position; obtaining a temperature stability probability at each moment of each position; obtaining a temperature monitoring threshold range at each moment of each position; and monitoring the temperature of a production line in a process in real time. The application aims to solve the problem that a static temperature standard cannot effectively monitor abnormalities in a high-temperature working process of a semiconductor production line, thereby affecting safety production.

Description

TECHNICAL FIELD

[0001] The application relates to the field of temperature measurement, and in particular to a semiconductor production line safety production management system based on Internet of Things. BACKGROUND

[0002] The processing and production of the semiconductor industry often need high-temperature operation, such as melting and diffusion processes, which need higher temperatures. If the operation is improper or the equipment fails, a fire is easily caused. Although some equipment has heat insulation design, local heat radiation is still brought during work, which causes the temperature change of the surrounding space.

[0003] The processing equipment and processing technology of different enterprises are different, and it is difficult to monitor whether the entire production line is abnormal by directly setting a fixed and single temperature threshold, that is, it is difficult to use a static judgment standard to detect whether the production line temperature is abnormal under different scenes and different processes, including the setting of a constant-temperature workshop for high-precision processes such as photolithography. Most enterprises or processes require that the overall temperature of the environment be stable and not fluctuate sharply. However, under the condition of overall temperature stability, there are still cases of local temperature being too high or too low. Not every temperature in the space is equal, which leads to the fact that the static fixed temperature standard cannot effectively monitor the temperature abnormality in the high-temperature process of the semiconductor production line in a timely manner, and seriously affects the safe production process of the semiconductor. SUMMARY

[0004] The application provides a semiconductor production line safety production management system based on Internet of Things to solve the problem that the existing static temperature standard cannot effectively monitor the abnormality in the high-temperature work process of the semiconductor production line and affects the safe production.

[0005] The application also provides a semiconductor production line safety production management system based on Internet of Things, which comprises:

[0006] A temperature data acquisition module is arranged to arrange sensors at multiple positions of the semiconductor production line, acquire temperature data at several time points of each position in multiple processes, and obtain a simultaneous time temperature set composed of temperature data of different processes at the same time at each position.

[0007] A temperature abnormality analysis module is arranged to obtain several approximate temperatures by rounding the temperature data in the simultaneous time temperature set, statistically acquire the occurrence probability of each approximate temperature, analyze the occurrence probability and performance of two types of approximate temperatures under each approximate temperature as a boundary in the simultaneous time temperature set, and analyze the mean performance of the approximate temperatures, calculate the boundary appropriateness of each approximate temperature by interclass variance, and further obtain the boundary temperature of the simultaneous time temperature set at each position at each time.

[0008] The temperature set at the same moment is divided into an upper temperature set and a lower temperature set based on the boundary temperature. The frequency of occurrence of each temperature data is combined with the cumulative frequency of occurrence and its proportion in the total frequency of occurrence of the upper temperature set or the lower temperature set to obtain the basic threshold ratio of the upper temperature set and the lower temperature set respectively. The high temperature degree at each location and time is obtained based on the difference between the temperature data in the temperature set at the same moment and the maximum value of the temperature data at the corresponding location.

[0009] Based on the difference in temperature data at adjacent moments and the simultaneous temperature set, the probability of temperature stability at each location and time is obtained. The upper and lower threshold adjustment coefficients for each location and time are obtained by combining the high temperature level at each location and time, as well as the basic threshold ratio of the upper and lower temperature sets. The temperature monitoring threshold range for each location and time is obtained based on the boundary temperature.

[0010] The abnormal prompt and visualization module is used to monitor the temperature of the production line in the semiconductor production line in real time according to the temperature monitoring threshold range at various locations of the production line, and to display the monitoring results and provide temperature abnormality prompts through visualization and buzzer.

[0011] Optionally, the occurrence probability of each approximate temperature is obtained by:

[0012] For a set of temperatures at any moment in any location, count the number of occurrences of any approximate temperature among all the temperature data in the temperature set at that moment, and use the ratio of the number of occurrences to the number of temperature data in the temperature set at that moment as the probability of occurrence of the approximate temperature.

[0013] Optionally, the specific method for obtaining the appropriateness of the boundaries of the approximate temperatures is as follows:

[0014] According to the approximate temperatures in the temperature set at each location and time, the high temperature set and the low temperature set of each approximate temperature are obtained, as well as the corresponding high temperature probability and low temperature probability;

[0015] For any position at any time in the temperature set The approximate temperature is calculated by adding the approximate temperature of each temperature data in the high temperature set as the first approximate temperature. The average temperature of the high approximate temperature; the cumulative value of the approximate temperature of each temperature data in the low temperature set is used as the first The lower average temperature of the approximate temperature; the cumulative value of the approximate temperature of all temperature data in the temperature set at the same moment is used as the average temperature of the location at that moment;

[0016] The position at this moment The appropriateness of the boundary of the approximate temperature The calculation method is:

[0017]

[0018] in, Indicates the The probability of a high temperature for an approximate temperature, Indicates the The average temperature of the approximate temperature is higher, Indicates the The probability of a low temperature for an approximate temperature, Indicates the The lower average temperature of the approximate temperature, Indicates the average temperature at that location at that moment.

[0019] Optionally, the specific method of obtaining the high-temperature set and the low-temperature set of each approximate temperature, and the corresponding high-temperature probability and low-temperature probability, includes:

[0020] For any position at any time in the temperature set An approximate temperature, The approximate temperature is the boundary, and the approximate temperature in the temperature set at the same moment is greater than the Several temperature data of approximate temperature constitute the The high temperature set of the approximate temperature is less than or equal to the Several temperature data of approximate temperature constitute the A collection of low-temperature classes with approximate temperatures;

[0021] The cumulative value of the occurrence probability of the approximate temperature of each temperature data in the high temperature set is used as the first The probability of the high temperature of the approximate temperature is calculated; the cumulative value of the occurrence probability of the approximate temperature of each temperature data in the low temperature set is used as the first The probability of a low temperature for an approximate temperature.

[0022] Optionally, the step of obtaining the boundary temperature of the set of temperatures at each position and at each time includes the following specific methods:

[0023] The approximate temperature corresponding to the maximum boundary suitability among the approximate temperatures in the simultaneous temperature set at any position at any time is used as the boundary temperature of the simultaneous temperature set at that position at that time.

[0024] Optionally, the basic threshold ratios of the upper limit temperature set and the lower limit temperature set are specifically obtained as follows:

[0025] Obtain the ratio of the number of times each temperature data in any temperature set at the same moment appears in the temperature set at the same moment to the total number of temperature data in the temperature set at the same moment as the frequency of occurrence of each temperature data;

[0026] For the upper limit temperature set, the sum of the occurrence frequencies of all temperature data in the upper limit temperature set is used as the total occurrence frequency of the upper limit temperature set. The sum of the occurrence frequencies of each temperature data is accumulated in the process of traversing from the minimum value of the temperature data, and the ratio is obtained by traversing the upper limit temperature set one by one from small to large. When the ratio is greater than the stop threshold for the first time, the traversal is stopped, and the currently traversed temperature data is used as the basic monitoring threshold of the upper limit temperature set. The basic threshold ratio of the upper limit temperature set is 1 / 4. The calculation method is:

[0027]

[0028] in, Indicates the upper temperature set basic monitoring threshold, Indicates the maximum value of the temperature data in the temperature set at that moment, Indicates the boundary temperature of the temperature set at that moment;

[0029] For the lower limit temperature set, obtain the total occurrence frequency of the lower limit temperature set, obtain the sum of the occurrence frequencies of each temperature data in the process of traversing from the maximum temperature data, and obtain the ratio with the total occurrence frequency of the lower limit temperature set. Traverse the temperature data one by one from large to small in the lower limit temperature set. When the ratio corresponding to the lower limit temperature set is greater than the stop threshold for the first time, stop traversing and use the currently traversed temperature data as the basic monitoring threshold of the lower limit temperature set. The basic threshold ratio of the lower limit temperature set is 0. The calculation method is:

[0030]

[0031] in, Indicates the lower limit temperature set basic monitoring threshold, Indicates the minimum value of the temperature data in the temperature set at that moment, Indicates the boundary temperature of the temperature set at that moment.

[0032] Optionally, the method of obtaining the high temperature level at each position at each time includes:

[0033]

[0034] in, Indicates any position The high temperature at that moment, Indicates the number of temperature data in the temperature set at the same moment, Indicates the position The temperature set at the same time Temperature data, Indicates the minimum value of the temperature data at all times of each processing at this position. Indicates the maximum value of the temperature data at all times of each processing at this position.

[0035] Optionally, the specific method of obtaining the temperature stability probability at each position and time is as follows:

[0036] For any position and any processing The temperature data at the moment is used to obtain its judgment window based on several adjacent temperature data; the mean of all temperature data in the judgment window is compared with the The absolute value of the temperature difference at the moment is used as the value of the next processing at this position. The degree of temperature change at the moment; if the degree of temperature change is greater than or equal to the temperature change threshold, The constant temperature state at the moment is marked as 0, recorded as the The constant temperature state value at the moment; if the temperature change degree is less than the temperature change threshold, The constant temperature state at the moment is marked as 1, and recorded as Constant temperature state value at a moment;

[0037] Process the position each time The average value of the constant temperature state at the moment is taken as the The probability of temperature stability at a certain moment.

[0038] Optionally, the obtaining of the upper threshold adjustment coefficient and the lower threshold adjustment coefficient at each position and at each time includes the following specific methods:

[0039]

[0040] in, Indicates the upper threshold adjustment coefficient at any location at any time, Indicates the basic threshold ratio of the upper temperature limit set at this location at this moment. represents the probability of temperature stability at this location at this moment, Indicates the high temperature level at this location at this moment; the lower threshold adjustment coefficient at this location at this moment The calculation method is:

[0041]

[0042] in, Indicates the basic threshold ratio of the lower limit temperature set at this location at this time.

[0043] Optionally, the specific method of obtaining the temperature monitoring threshold range at each location and time includes:

[0044] For any location at any time, the temperature monitoring threshold range is obtained as follows:

[0045]

[0046]

[0047] in, Indicates the upper limit of the temperature monitoring threshold range at this location at this moment. Indicates the lower limit of the temperature monitoring threshold range at this location at this moment. Indicates the upper threshold adjustment coefficient of the position at this moment, Indicates the lower threshold adjustment coefficient of the position at this moment, Indicates the boundary temperature of the temperature set at this location at this moment, Indicates the maximum value of the temperature data in the temperature set at this location at this moment. Indicates the minimum value of the temperature data in the temperature set at the same time at this location;

[0048] The closed interval formed by the lower limit and the upper limit of the temperature monitoring threshold range is used as the temperature monitoring threshold range at the position at the time.

[0049] The beneficial effects of the present invention are as follows: the present invention deploys temperature sensors at multiple locations of a semiconductor production line, monitors temperature data in real time during multiple processing processes to construct a simultaneous temperature set; based on the maximum inter-class variance idea and combined with the approximate performance of temperature data, after obtaining the boundary temperature for the simultaneous temperature set, the boundary temperature is used to divide the upper and lower temperature sets, which are respectively used to analyze the upper and lower limits of the temperature monitoring thresholds of the corresponding positions at the corresponding moments under each processing; and for the upper and lower temperature sets, based on the performance of the temperature data therein, the main distribution of the vast majority of temperature data therein is analyzed, thereby obtaining the basic monitoring threshold and the basic threshold ratio, which are used for adjustment based on a large amount of temperature data in the subsequent upper and lower limit adjustment process of the monitoring threshold; At the same time, based on the changes in the time series data within the same processing, the high temperature degree and constant temperature stage are analyzed, and the range of the monitoring threshold is adjusted according to the processing state at the corresponding moment, and finally the temperature monitoring threshold range is obtained; based on the analysis of the temperature set and time series data at the same moment, the main distribution of temperature data in the temperature set at the same moment, as well as the high temperature performance, are quantified. At the same time, combined with the constant temperature performance of the temperature data under the time series data, the temperature monitoring threshold range is finally adjusted at each moment to realize the dynamic range monitoring of the temperature at each position of the production line during the processing process, and to perform visual display and abnormal prompts, so as to timely discover temperature anomalies in the production line through sensor network and Internet of Things data analysis to ensure the safe production process of semiconductors. BRIEF DESCRIPTION OF THE DRAWINGS

[0050] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the following briefly introduces the drawings required for use in the embodiments or the description of the prior art. 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 any creative work.

[0051] Figure 1 A structural block diagram of a semiconductor production line safety production management system based on the Internet of Things provided by one embodiment of the present invention. DETAILED DESCRIPTION

[0052] The following will clearly and completely describe the technical solutions in the embodiments of the present invention in conjunction with the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making creative efforts are within the scope of protection of the present invention.

[0053] See also Figure 1 , which shows a structural block diagram of a semiconductor production line safety production management system based on the Internet of Things provided by one embodiment of the present invention, the system includes:

[0054] The temperature data acquisition module 101 arranges sensors at multiple locations of the semiconductor production line to collect temperature data of each location at several moments during multiple processing, and obtains a simultaneous temperature set consisting of temperature data of different processing times at the same moment at each location.

[0055] The purpose of this embodiment is to solve the problem of abnormality monitoring in uncertain scenarios. That is, the semiconductor industry has different polysilicon processing technologies and process temperature requirements. Humans can only provide rough judgment standards or require exactly the same process parameters. At the same time, the temperature of the entire production line does not fluctuate drastically, but the temperature of a part of the equipment may be too high, and the temperature around the cooling equipment may be too low. The temperature at each location may fluctuate, and local anomalies are difficult to identify.

[0056] The scenario targeted by this embodiment is the semiconductor industry's production line for polysilicon processing. Temperature sensors are arranged on the production line to detect real-time temperature, and data is uploaded to the network through Internet of Things technology. The data is processed on the server to analyze whether the temperature data is abnormal, and abnormalities are discovered in time to achieve safe production management.

[0057] It should be noted that since the different processing processes of the semiconductor production line are the same, that is, the production process is the same, the processing time of multiple processing is the same, and the temperature sensor starts collecting data at the beginning of a single processing and ends collecting data after the processing is completed, the temperature data can be time-aligned in different processing times.

[0058] Specifically, thermocouple sensors or near-infrared sensors are used to deploy sensors at multiple locations on the semiconductor production line and monitor the temperature. In this embodiment, the sampling time interval is set to 10 seconds, and temperature data of each position at each moment of multiple processing is obtained; and since the sampling time interval is the same and the processing time is the same, the amount of temperature data of the same position in different processing is the same, and the temperature data of the same position in each processing are sorted row by row and column by column according to the order value. The same row is the temperature data of the same processing, and the same column is the temperature data of different processing with the same order value. The temperature data in the same column constitute a simultaneous temperature set, and the first temperature data in each processing corresponds to the first moment, so each simultaneous temperature set corresponds to a moment of each processing, and a simultaneous temperature set of each moment at the same position is obtained.

[0059] Temperature anomaly analysis module 102:

[0060] (1) The temperature data in the temperature set at the same moment are rounded off to obtain several approximate temperatures, and the probability of occurrence of each approximate temperature is statistically obtained; the probability of occurrence and performance of the two types of approximate temperatures under the boundary of each approximate temperature in the temperature set at the same moment and the performance of the approximate temperature mean are analyzed, and the degree of suitability of the boundary of each approximate temperature is obtained by calculating the inter-class variance, and then the boundary temperature of the temperature set at each position and time is obtained.

[0061] Preferably, in one embodiment of the present invention, the temperature data in the temperature set at the same moment are rounded off to obtain several approximate temperatures, and the occurrence probability of each approximate temperature is statistically obtained, including the specific method of:

[0062] It should be noted that when the same temperature data or similar temperature data exist in the temperature set at the same time, the approximate temperature is obtained through rounding approximation operation, which provides a basis for the subsequent between-class variance calculation.

[0063] Specifically, for a simultaneous temperature set at any position at any time, all temperature data in the simultaneous temperature set are rounded to obtain the approximate temperature of each temperature data; the number of occurrences of any approximate temperature in the approximate temperatures of all temperature data in the simultaneous temperature set is counted, and the ratio of the number of occurrences to the number of temperature data in the simultaneous temperature set (total number of processing times) is used as the probability of occurrence of the approximate temperature.

[0064] It should be further explained that, in the simultaneous temperature set, it is necessary to obtain the boundary temperature based on the approximate temperature in order to divide the simultaneous temperature set into two categories of high and low temperatures. Considering the maximum inter-class variance process, the probability of occurrence and mean performance of the approximate temperature in the two categories divided under each approximate temperature are used as the boundary. Referring to the maximum inter-class variance idea, the greater the difference between the two categories of temperature, the more appropriate the boundary is, thereby quantifying the degree of suitability and obtaining the boundary temperature.

[0065] Preferably, in one embodiment of the present invention, the probability of occurrence and performance of the two types of approximate temperatures under the boundary of each approximate temperature in the simultaneous temperature set and the performance of the approximate temperature mean are analyzed, and the appropriateness of the boundary of each approximate temperature is obtained by calculating the inter-class variance, thereby obtaining the boundary temperature of the simultaneous temperature set at each position and time. The specific method includes:

[0066] For any position at any time in the temperature set An approximate temperature, The approximate temperature is the boundary, and the approximate temperature in the temperature set at the same moment is greater than the Several temperature data of approximate temperature constitute the The high temperature set of the approximate temperature is less than or equal to the Several temperature data of approximate temperature constitute the The low-class temperature set of approximate temperatures; the cumulative value of the occurrence probability of the approximate temperature of each temperature data in the high-class temperature set (when the approximate temperatures of different temperature data are the same, it is only calculated once, that is, the statistical probability accumulation is performed on the approximate temperature) is used as the first Similarly, the probability of the approximate temperature of each temperature data in the low temperature set is accumulated as the probability of the high temperature of the approximate temperature of the first approximate temperature; The probability of a low temperature for an approximate temperature.

[0067] Furthermore, the accumulated value of the approximate temperature of each temperature data in the high temperature set (when the approximate temperatures of different temperature data are the same, the temperature data is used as the benchmark and all need to be accumulated, that is, the corresponding approximate temperature accumulation is performed for the temperature data, and the same temperature data also needs to be accumulated multiple times, and the accumulation is not for the temperature data value, but for the temperature data itself) is used as the first Similarly, the cumulative value of the approximate temperature of each temperature data in the low temperature set is used as the average temperature of the approximate temperature of the first approximate temperature; The lower average temperature of the approximate temperature of the temperature data in the temperature set at the same moment, that is, the sum of the approximate temperatures of each temperature data, is taken as the average temperature of the position at that moment; then the first The appropriateness of the boundary of the approximate temperature The calculation method is:

[0068]

[0069] in, Indicates the The probability of a high temperature for an approximate temperature, Indicates the The average temperature of the approximate temperature is higher, Indicates the The probability of a low temperature for an approximate temperature, Indicates the The lower average temperature of the approximate temperature, Indicates the average temperature at that location at that moment.

[0070] It should be noted that the sum of the occurrence probabilities of the two types of temperatures is used as the weight of the sum of squares of the differences between the corresponding average temperature and the overall average temperature, so as to obtain the between-class variance based on the approximate temperature, which also reflects the suitability of the corresponding approximate temperature as a dividing line.

[0071] Furthermore, the approximate temperature corresponding to the maximum boundary suitability among the approximate temperatures in the temperature set at the location at the moment is used as the boundary temperature of the temperature set at the location at the moment.

[0072] (2) The temperature set at the same moment is divided into an upper limit temperature set and a lower limit temperature set according to the boundary temperature. Combined with the frequency of occurrence of each temperature data, the basic threshold ratio of the upper limit temperature set and the lower limit temperature set is obtained according to the continuous accumulation of the frequency of occurrence and the proportion in the total frequency of occurrence of the upper limit temperature set or the lower limit temperature set; based on the difference between the temperature data in the temperature set at the same moment and the maximum value of the temperature data at the corresponding position, the high temperature degree at each position and at each moment is obtained.

[0073] Preferably, in one embodiment of the present invention, the temperature set at the same moment is divided into an upper limit temperature set and a lower limit temperature set according to the boundary temperature. In combination with the occurrence frequency of each temperature data, the basic threshold ratio of the upper limit temperature set and the lower limit temperature set is obtained based on the continuous accumulation of the occurrence frequency and the proportion of the occurrence frequency in the total occurrence frequency of the upper limit temperature set or the lower limit temperature set, respectively. The specific method includes:

[0074] For a simultaneous temperature set at any position at any moment, the simultaneous temperature set is divided into an upper limit temperature set and a lower limit temperature set based on the boundary temperature of the simultaneous temperature set, wherein the upper limit temperature set includes all temperature data in the simultaneous temperature set that is greater than the boundary temperature, and the lower limit temperature set includes all temperature data in the simultaneous temperature set that is less than or equal to the boundary temperature.

[0075] Furthermore, the ratio of the number of times each temperature data in the temperature set at the same moment appears in the temperature set at the same moment to the total number of temperature data in the temperature set at the same moment is obtained as the frequency of occurrence of each temperature data; a preset stop threshold is set, and the stop threshold in this embodiment is described as 0.9; for the upper limit temperature set, the sum of the occurrence frequencies of all temperature data in the upper limit temperature set is used as the total frequency of occurrence of the upper limit temperature set, and the sum obtained by accumulating the occurrence frequencies of each temperature data in the process of traversing from the minimum value of the temperature data is obtained, and the ratio is obtained by traversing the upper limit temperature set one by one from small to large. When the ratio is greater than the stop threshold for the first time, the traversal is stopped, and the currently traversed temperature data is used as the basic monitoring threshold of the upper limit temperature set. Then, the basic threshold ratio of the upper limit temperature set is 0. The calculation method is:

[0076]

[0077] in, Indicates the upper temperature set basic monitoring threshold, Indicates the maximum value of the temperature data in the temperature set at that moment, Indicates the boundary temperature of the temperature set at that moment.

[0078] Furthermore, for the lower limit temperature set, the total occurrence frequency of the lower limit temperature set is obtained in the same way, and the sum of the occurrence frequencies of each temperature data is accumulated in the process of traversing from the maximum value of the temperature data, and the ratio is obtained by traversing the lower limit temperature set one by one from large to small. When the ratio is greater than the stop threshold for the first time, the traversal is stopped, and the currently traversed temperature data is used as the basic monitoring threshold of the lower limit temperature set. The basic threshold ratio of the lower limit temperature set is 0. The calculation method is:

[0079]

[0080] in, Indicates the lower limit temperature set basic monitoring threshold, Indicates the minimum value of the temperature data in the temperature set at that moment, Indicates the boundary temperature of the temperature set at that moment.

[0081] It should be noted that the basic threshold ratio reflects the distribution range of a large amount of temperature data in the corresponding upper and lower limit temperature sets, so as to obtain the basic monitoring threshold, and quantify the proportion based on the monitoring threshold and the boundary temperature to provide a basis for the subsequent upper and lower limit threshold adjustment coefficients.

[0082] Preferably, in one embodiment of the present invention, the high temperature degree at each position at each moment is obtained based on the difference between the temperature data in the temperature set at the same moment and the maximum value of the temperature data at the corresponding position, including the specific method of:

[0083] For any position At this moment, the position High temperature at the moment The calculation method is:

[0084]

[0085] in, Indicates the number of temperature data in the temperature set at the same moment, Indicates that the position The temperature set at the same time Temperature data, Indicates the minimum value of the temperature data at all times of each processing at this position. Indicates the maximum value of the temperature data at all times of each processing at this position.

[0086] It should be noted that the temperature data in the temperature set at the same time are normalized by the maximum and minimum values, and are closer to the maximum value of the temperature data at all processing moments at that position. The temperature data at that moment is generally biased towards high temperature, thereby quantifying the degree of high temperature.

[0087] (3) Based on the difference in temperature data at adjacent moments and the temperature set at the same moment, the probability of temperature stability at each location at each moment is obtained; combined with the high temperature level at each location at each moment and the basic threshold ratio of the upper limit temperature set and the lower limit temperature, the upper limit threshold adjustment coefficient and the lower limit threshold adjustment coefficient at each location at each moment are obtained; based on the boundary temperature, the temperature monitoring threshold range at each location at each moment is obtained.

[0088] It should be noted that similar diffusion processes on the production line can be divided into three stages: heating, constant temperature (diffusion reaction), and cooling. Similarly, processes that require heating can be roughly divided into the above three stages. For example, in the cutting process, the temperature increases during the process of increasing the rotation speed. The constant temperature stage is the main stage for achieving the process effect and is also the stage that requires intensive monitoring. The temperature in this stage is generally high, and there is a certain high temperature risk. The data needs to be analyzed, and the monitoring threshold is calculated and adjusted according to the temperature stability. (The more accurate the monitoring, the earlier the abnormality can be discovered, otherwise the loss will be increased. The abnormality is caused by improper operation or equipment failure.)

[0089] Preferably, in one embodiment of the present invention, the temperature stability probability of each position at each time is obtained based on the difference performance of temperature data at adjacent times and the temperature set at the same time, including the specific method of:

[0090] For any position and any processing The temperature data at the moment, the present embodiment presets the judgment window as 4 for description, and the position is processed in the same time. The temperature data of the four adjacent moments after the moment and the The temperature data at the moment is used as the The judgment window of the temperature data at each moment is specially explained. If the number of moments with temperature data after any moment is less than the judgment window, the temperature data of the adjacent moments before the temperature data at that moment will be selected one by one until the number of judgment window construction is met; the mean of all temperature data in the judgment window is compared with the first The absolute value of the temperature difference at the moment is used as the value of the next processing at this position. The temperature change degree at each moment; preset temperature change threshold, the temperature change threshold in this embodiment is described as 0.5, if the temperature change degree is greater than or equal to the temperature change threshold, the first The constant temperature state at the moment is marked as 0, recorded as the The constant temperature state value at the moment; if the temperature change degree is less than the temperature change threshold, The constant temperature state at the moment is marked as 1, and recorded as The constant temperature state value at each moment; the temperature in the diffusion process is in the range of 800-1100℃, and the allowed fluctuation degree is .

[0091] Further, obtain the position of each processing The constant temperature state value at the moment, that is, the The constant temperature state value of each temperature data in the temperature set at the same moment, the mean of all constant temperature state values ​​in the temperature set at the same moment, is taken as the first value of the position. The probability of temperature stability at a certain moment.

[0092] What needs to be explained is that the difference in continuous temperature data in the time series data under the same position and the same processing is combined with the threshold to judge the quantified constant temperature state, and the temperature stability probability is quantified by the constant temperature state in the temperature set at the same moment. The more constant temperature state values ​​1, the more temperature data are at a constant temperature, the greater the probability of temperature stability, and the smaller the subsequent threshold adjustment coefficient needs to be.

[0093] Preferably, in one embodiment of the present invention, the upper threshold adjustment coefficient and the lower threshold adjustment coefficient at each position and at each time are obtained by combining the high temperature degree at each position and at each time, and the basic threshold ratio of the upper temperature set and the lower temperature set. The specific method includes:

[0094] For any position at any time, the upper threshold adjustment coefficient The calculation method is:

[0095]

[0096] in, Indicates the basic threshold ratio of the upper temperature limit set at this location at this moment. represents the probability of temperature stability at this location at this moment, Indicates the high temperature level at that location at that moment.

[0097] Similarly, the lower threshold adjustment coefficient of this position at this moment The calculation method is:

[0098]

[0099] in, Indicates the basic threshold ratio of the lower limit temperature set at this location at this time. represents the probability of temperature stability at this location at this moment, Indicates the high temperature level at that location at that moment.

[0100] What needs to be explained is that based on the basic threshold ratio, the threshold coefficient is adjusted in combination with the high temperature degree and the constant temperature stage. The greater the high temperature degree, the greater the probability of temperature stability. In the continuous high temperature constant temperature stage, the temperature performance needs to be more stable, and the smaller the threshold monitoring range, the smaller the threshold adjustment coefficient.

[0101] Preferably, in one embodiment of the present invention, the temperature monitoring threshold range at each position and time is obtained based on the boundary temperature, including the following specific methods:

[0102] For any location at any time, the temperature monitoring threshold range is obtained as follows:

[0103]

[0104]

[0105] in, Indicates the upper limit of the temperature monitoring threshold range at this location at this moment. Indicates the lower limit of the temperature monitoring threshold range at this location at this moment. Indicates the upper threshold adjustment coefficient of the position at this moment, Indicates the lower threshold adjustment coefficient of the position at this moment, Indicates the boundary temperature of the temperature set at this location at this moment, Indicates the maximum value of the temperature data in the temperature set at this location at this moment. Indicates the minimum temperature data value in the temperature set at this location at this moment.

[0106] Furthermore, a closed interval formed by the lower limit and the upper limit of the temperature monitoring threshold range is used as the temperature monitoring threshold range at the position at the time.

[0107] It should be noted that the smaller the threshold adjustment coefficient is, the smaller the portion adjusted based on the boundary temperature in combination with the maximum and minimum values ​​of the temperature data is, thereby obtaining the temperature monitoring threshold range.

[0108] At this point, after obtaining the boundary temperature for the simultaneous temperature set based on the maximum inter-class variance idea and combined with the approximate performance of the temperature data, the upper and lower limit temperature sets are divided according to the boundary temperature, which are used to analyze the upper and lower limits of the temperature monitoring threshold of the corresponding position at the corresponding moment under each processing; and for the upper and lower limit temperature sets, based on the performance of the temperature data therein, the main distribution of the vast majority of temperature data in them is analyzed to obtain the basic monitoring threshold and the basic threshold ratio, which is used to adjust the upper and lower limits of the monitoring threshold based on a large amount of temperature data in the subsequent adjustment process; at the same time, according to the changes in the time series data within the same processing, the high temperature degree and constant temperature stage are analyzed, and according to the processing state at the corresponding moment, the range of the monitoring threshold is adjusted to finally obtain the temperature monitoring threshold range.

[0109] The abnormal prompt and visualization module 103 monitors the temperature of the production line in the process in real time according to the temperature monitoring threshold range at each position of the semiconductor production line, and displays the monitoring results and temperature abnormality prompts through visualization and buzzer.

[0110] It should be noted that after obtaining the temperature monitoring threshold range for each position at each moment, the temperature monitoring threshold range for consecutive moments at the same position actually constitutes a curve that cannot be exceeded and a curve that cannot be lower than in the time series of the temperature data of the position, and the two curves are obtained based on several complete processing processes in the past. The production line temperature can be monitored in real time based on the corresponding curves, and an abnormal prompt can be issued in time after exceeding the temperature monitoring threshold range.

[0111] Specifically, for the current processing process at any position, each moment of the current processing has been aligned with the moments of multiple historical processing through timing alignment. The temperature monitoring threshold range corresponding to each moment of the current processing at this position is transmitted to the ESP32 chip in real time via WiFi, and the monitoring threshold range and real-time temperature data at that moment are displayed on the OLED screen. If the temperature data at any moment is not within the temperature monitoring threshold range at that moment, the temperature data is abnormal, and the abnormal signal is transmitted to the ESP32 chip to control the LED light to flash and the buzzer to alarm, thereby realizing temperature abnormality prompt.

[0112] At this point, by placing temperature sensors at multiple locations on the semiconductor production line, real-time monitoring of temperature data during multiple processing processes is used to construct a simultaneous temperature set. Based on the analysis of the simultaneous temperature set and time series data, the main distribution of temperature data in the simultaneous temperature set and high-temperature performance are quantified. At the same time, combined with the constant temperature performance of temperature data under time series data, the temperature monitoring threshold range is finally adjusted at each moment to achieve dynamic range monitoring of temperature at various locations on the production line during the processing process, and visual display and abnormal prompts are performed. Through sensor network and Internet of Things data analysis, temperature anomalies in the production line can be discovered in a timely manner to ensure the safe production process of semiconductors.

[0113] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc. made within the principles of the present invention should be included in the scope of protection of the present invention.

Claims

1. A semiconductor production line safety management system based on the Internet of Things, characterized by: The system includes: The temperature data acquisition module is used to arrange sensors at multiple locations in the semiconductor production line to collect temperature data at multiple processing locations at several times, and obtain a simultaneous temperature set consisting of temperature data of different processing times at the same time at each location; The temperature anomaly analysis module is used to round off the temperature data in the simultaneous temperature set to obtain several approximate temperatures and statistically obtain the probability of occurrence of each approximate temperature. It also analyzes the probability of occurrence and performance of two approximate temperature categories under each approximate temperature set as the boundary, as well as the performance of the approximate temperature mean. It calculates the inter-class variance to determine the appropriateness of the boundary of each approximate temperature, and then obtains the boundary temperature of the simultaneous temperature set at each location and time. The temperature set at the same moment is divided into an upper temperature set and a lower temperature set based on the boundary temperature. The frequency of occurrence of each temperature data is combined with the cumulative frequency of occurrence and its proportion in the total frequency of occurrence of the upper temperature set or the lower temperature set to obtain the basic threshold ratio of the upper temperature set and the lower temperature set respectively. The high temperature degree at each location and time is obtained based on the difference between the temperature data in the temperature set at the same moment and the maximum value of the temperature data at the corresponding location. Based on the difference in temperature data at adjacent moments and the simultaneous temperature set, the probability of temperature stability at each location and time is obtained. The upper and lower threshold adjustment coefficients for each location and time are obtained by combining the high temperature level at each location and time, as well as the basic threshold ratio of the upper and lower temperature sets. The temperature monitoring threshold range for each location and time is obtained based on the boundary temperature. The abnormal prompt and visualization module is used to monitor the temperature of the production line in the semiconductor production line in real time according to the temperature monitoring threshold range at various locations of the production line, and to display the monitoring results and provide temperature abnormality prompts through visualization and buzzer.

2. The semiconductor production line safety management system based on the Internet of Things according to claim 1, characterized in that: The specific method for obtaining the occurrence probability of each approximate temperature is as follows: For a set of temperatures at any moment in any location, count the number of occurrences of any approximate temperature among all the temperature data in the temperature set at that moment, and use the ratio of the number of occurrences to the number of temperature data in the temperature set at that moment as the probability of occurrence of the approximate temperature.

3. The semiconductor production line safety management system based on the Internet of Things according to claim 1, characterized in that: The specific method for obtaining the appropriateness of the boundaries of the approximate temperatures is as follows: According to the approximate temperatures in the temperature set at each location and time, the high temperature set and the low temperature set of each approximate temperature are obtained, as well as the corresponding high temperature probability and low temperature probability; For any position at any time in the temperature set The approximate temperature is calculated by adding the approximate temperature of each temperature data in the high temperature set as the first approximate temperature. The average temperature of the high approximate temperature; the cumulative value of the approximate temperature of each temperature data in the low temperature set is used as the first The lower average temperature of the approximate temperature; the cumulative value of the approximate temperature of all temperature data in the temperature set at the same moment is used as the average temperature of the location at that moment; The position at this moment The appropriateness of the boundary of the approximate temperature The calculation method is: in, Indicates the The probability of a high temperature for an approximate temperature, Indicates the The average temperature of the approximate temperature is higher, Indicates the The probability of a low temperature for an approximate temperature, Indicates the The lower average temperature of the approximate temperature, Indicates the average temperature at that location at that moment.

4. The semiconductor production line safety management system based on the Internet of Things according to claim 3, characterized in that: The specific method of obtaining the high-temperature set and the low-temperature set of each approximate temperature, as well as the corresponding high-temperature probability and low-temperature probability, includes: For any position at any time in the temperature set An approximate temperature, The approximate temperature is the boundary, and the approximate temperature in the temperature set at the same moment is greater than the Several temperature data of approximate temperature constitute the The high temperature set of the approximate temperature is less than or equal to the Several temperature data of approximate temperature constitute the A collection of low-temperature classes with approximate temperatures; The cumulative value of the occurrence probability of the approximate temperature of each temperature data in the high temperature set is used as the first The probability of the high temperature of the approximate temperature is calculated; the cumulative value of the occurrence probability of the approximate temperature of each temperature data in the low temperature set is used as the first The probability of a low temperature for an approximate temperature.

5. The semiconductor production line safety management system based on the Internet of Things according to claim 1, characterized in that: The specific method of obtaining the boundary temperature of the temperature set at each position and at each time includes: The approximate temperature corresponding to the maximum boundary suitability among the approximate temperatures in the simultaneous temperature set at any position at any time is used as the boundary temperature of the simultaneous temperature set at that position at that time.

6. The semiconductor production line safety management system based on the Internet of Things according to claim 1, characterized in that: The basic threshold ratios of the upper limit temperature set and the lower limit temperature set are obtained in the following way: Obtain the ratio of the number of times each temperature data in any temperature set at the same moment appears in the temperature set at the same moment to the total number of temperature data in the temperature set at the same moment as the frequency of occurrence of each temperature data; For the upper limit temperature set, the sum of the occurrence frequencies of all temperature data in the upper limit temperature set is used as the total occurrence frequency of the upper limit temperature set. The sum of the occurrence frequencies of each temperature data is accumulated in the process of traversing from the minimum value of the temperature data, and the ratio is obtained by traversing the upper limit temperature set one by one from small to large. When the ratio is greater than the stop threshold for the first time, the traversal is stopped, and the currently traversed temperature data is used as the basic monitoring threshold of the upper limit temperature set. The basic threshold ratio of the upper limit temperature set is 1 / 4. The calculation method is: in, Indicates the upper temperature set basic monitoring threshold, Indicates the maximum value of the temperature data in the temperature set at that moment, Indicates the boundary temperature of the temperature set at that moment; For the lower limit temperature set, obtain the total occurrence frequency of the lower limit temperature set, obtain the sum of the occurrence frequencies of each temperature data in the process of traversing from the maximum temperature data, and obtain the ratio with the total occurrence frequency of the lower limit temperature set. Traverse the temperature data one by one from large to small in the lower limit temperature set. When the ratio corresponding to the lower limit temperature set is greater than the stop threshold for the first time, stop traversing and use the currently traversed temperature data as the basic monitoring threshold of the lower limit temperature set. The basic threshold ratio of the lower limit temperature set is 0. The calculation method is: in, Indicates the lower limit temperature set basic monitoring threshold, Indicates the minimum value of the temperature data in the temperature set at that moment, Indicates the boundary temperature of the temperature set at that moment.

7. The semiconductor production line safety management system based on the Internet of Things according to claim 1, characterized in that: The specific method for obtaining the high temperature level at each position at each time is as follows: in, Indicates any position The high temperature at that moment, Indicates the number of temperature data in the temperature set at the same moment, Indicates that the position The temperature set at the same time Temperature data, Indicates the minimum value of the temperature data at all times of each processing at this position. Indicates the maximum value of the temperature data at all times of each processing at this position.

8. The semiconductor production line safety management system based on the Internet of Things according to claim 1, characterized in that: The specific method for obtaining the temperature stability probability at each position and time is as follows: For any position and any processing The temperature data at the moment is used to obtain its judgment window based on several adjacent temperature data; the mean of all temperature data in the judgment window is compared with the The absolute value of the temperature difference at the moment is used as the value of the next processing at this position. The degree of temperature change at the moment; if the degree of temperature change is greater than or equal to the temperature change threshold, The constant temperature state at the moment is marked as 0, recorded as the The constant temperature state value at the moment; if the temperature change degree is less than the temperature change threshold, The constant temperature state at the moment is marked as 1, and recorded as Constant temperature state value at a moment; Process the position each time The average value of the constant temperature state at the moment is taken as the The probability of temperature stability at a certain moment.

9. The semiconductor production line safety management system based on the Internet of Things according to claim 1, characterized in that: The specific method for obtaining the upper threshold adjustment coefficient and the lower threshold adjustment coefficient at each position and time includes: in, Indicates the upper threshold adjustment coefficient at any location at any time, Indicates the basic threshold ratio of the upper temperature limit set at this location at this moment. represents the probability of temperature stability at this location at this moment, Indicates the high temperature level at this location at this moment; the lower threshold adjustment coefficient at this location at this moment The calculation method is: in, Indicates the basic threshold ratio of the lower limit temperature set at this location at this time.

10. The semiconductor production line safety management system based on the Internet of Things according to claim 1, characterized in that: The specific method of obtaining the temperature monitoring threshold range at each location and time includes: For any location at any time, the temperature monitoring threshold range is obtained as follows: in, Indicates the upper limit of the temperature monitoring threshold range at this location at this moment. Indicates the lower limit of the temperature monitoring threshold range at this location at this moment. Indicates the upper threshold adjustment coefficient of the position at this moment, Indicates the lower threshold adjustment coefficient of the position at this moment, Indicates the boundary temperature of the temperature set at this location at this moment, Indicates the maximum value of the temperature data in the temperature set at this location at this moment. Indicates the minimum value of the temperature data in the temperature set at the same time at this location; The closed interval formed by the lower limit and the upper limit of the temperature monitoring threshold range is used as the temperature monitoring threshold range at the position at the time.

Images