Anomaly detection system

Abstract

In an abnormality detection system, there are provided: a waveform acquisition unit (61) that acquires waveform data that changes over time; and an abnormality degree calculation unit (62) that calculates an abnormality degree based on a cumulative value obtained by cumulating absolute values of differences in data values per unit time based on the waveform data acquired by the waveform acquisition unit. There is also provided an abnormality determination unit (64) that determines whether the waveform data is normal or abnormal based on the abnormality degree calculated by the abnormality degree calculation unit.

Description

[0001] Cross Reference to Related Applications

[0002] This application is based on Japanese Patent Application No. 2019-146524 filed on August 8, 2019, the content of which is incorporated herein by reference. TECHNICAL FIELD

[0003] The present application relates to an abnormality detection system that detects a problem device, a problem product, or the like using waveform data. BACKGROUND

[0004] When forming a resist mask on a semiconductor wafer, a coating process of coating a resist on a surface of the semiconductor wafer, an exposure process, a development process, and the like are performed. For example, in the coating process, resist coating, a heating process, and a cooling process are performed, and in the development process, a process of performing a heating process and a cooling process for development is performed. By placing the semiconductor wafer on a hot plate and putting it into a heating device maintained at a desired heating temperature in each heating process, and placing the semiconductor wafer on a cooling plate and putting it into a cooling device maintained at a desired cooling temperature in the cooling process, the semiconductor wafer is controlled to a desired temperature.

[0005] At this time, in order to suppress the occurrence of a problem of the semiconductor wafer to a minimum, it is necessary to be able to more quickly and correctly determine whether the semiconductor wafer is correctly placed on the hot plate or the cooling plate and is normally processed.

[0006] In a case where the semiconductor wafer is transported with a position offset, or the semiconductor wafer is transported with being placed on a guide provided in the hot plate or the cooling plate, a heating process and a cooling process of the semiconductor wafer are performed in this state, and a problem occurs. Such a problem is found in an inspection process as a subsequent process, but when the problem is detected in the inspection process, a large number of problem products have been manufactured in many cases. Therefore, it is desirable to detect the occurrence of a problem at the same time as the end of the processing of the coating process and the development process of the resist.

[0007] Therefore, in the coating process and the development process of the resist, a position offset is detected based on the temperature of the hot plate and the cooling plate, and a wafer that is normally processed without a position offset and a wafer that is abnormally processed with a position offset are discriminated.

[0008] For example, the semiconductor wafer heated in the heating process is transported and placed on the cooling plate, and since it is heated to 100 to 200°C in the previous heating process, heat is imparted to the cooling plate from the semiconductor wafer having a high temperature. Therefore, the temperature of the cooling plate temporarily rises. However, if the transport of the semiconductor wafer is not reliably performed, the semiconductor wafer is placed on the guide portion and is placed on the cooling plate in a tilted state, the distance between the cooling plate and the semiconductor wafer is increased and heat conduction is reduced, and the temperature rise of the cooling plate is small. The same situation occurs when the semiconductor wafer is heated while being placed on the hot plate. Using this phenomenon, in Patent Document 1, the temperature curve of the hot plate in the case of the time change and the temperature cumulative area of the range surrounded by the set temperature are calculated using a graph, and they are compared with a threshold value to perform abnormality detection.

[0009] Prior Art Documents

[0010] Patent Documents

[0011] Patent Document 1: Japanese Patent Application Laid-Open No. 2002-50557 SUMMARY

[0012] However, in the abnormality detection method of Patent Document 1, it is confirmed that there is a case where the temperature cumulative area of normal data, which is data in a normal processing case, and the temperature cumulative area of abnormal data, which is data in an abnormal processing case, do not differ. Therefore, it is not possible to reliably discriminate between a wafer in normal processing and a wafer in abnormal processing, and a wafer in normal processing can be erroneously detected as a wafer in abnormal processing. In such a case, even if normal processing is performed, a manufacturing interruption or the like must be performed, which hinders production.

[0013] An object of the present application is to provide an abnormality detection system capable of more rapidly and correctly performing abnormality detection.

[0014] An abnormality detection system according to one aspect of the present application includes a waveform acquisition section that acquires waveform data that changes in time series, an abnormality degree calculation section that calculates an abnormality degree based on a cumulative value obtained by accumulating the absolute value of the difference in data value per unit time based on the waveform data acquired by the waveform acquisition section, and an abnormality determination section that determines whether the waveform data is normal or abnormal based on the abnormality degree calculated by the abnormality degree calculation section.

[0015] Thus, the cumulative value obtained by accumulating the absolute value of the difference in data value per unit time is found based on the waveform data that changes in time series, and the abnormality degree is calculated based on the cumulative value. Further, based on the calculated abnormality degree, it is determined whether the waveform data is normal or abnormal. Therefore, correct abnormality detection can be performed, and the abnormality detection can be performed each time, so the abnormality detection can be performed more rapidly.

[0016] In addition, the parenthesized markings assigned to each constituent element indicate an example of the correspondence between that constituent element and the specific constituent elements described in the embodiments described later. Attached Figure Description

[0017] Figure 1A This is a perspective view showing the formation process of the resist mask using the coating / developing apparatus as described in the first embodiment.

[0018] Figure 1B It means to continue Figure 1A A three-dimensional diagram of the process of forming a resist mask.

[0019] Figure 1C It means to continue Figure 1B A three-dimensional diagram of the process of forming a resist mask.

[0020] Figure 1D It means to continue Figure 1C A three-dimensional diagram of the process of forming a resist mask.

[0021] Figure 1E It means to continue Figure 1D A three-dimensional diagram of the process of forming a resist mask.

[0022] Figure 1F It means to continue Figure 1E A three-dimensional diagram of the process of forming a resist mask.

[0023] Figure 1G It means to continue Figure 1F A three-dimensional diagram of the process of forming a resist mask.

[0024] Figure 1H It means to continue Figure 1G A three-dimensional diagram of the process of forming a resist mask.

[0025] Figure 1I It means to continue Figure 1H A three-dimensional diagram of the process of forming a resist mask.

[0026] Figure 2 This is a diagram representing the modular structure of an anomaly detection system.

[0027] Figure 3A It is a graph showing the temperature change of the hot plate over time during normal processing.

[0028] Figure 3B This is a graph showing the temperature change of the hot plate over time during abnormal handling.

[0029] Figure 4A This is a graph showing the temperature change of the cooling plate over time during normal processing.

[0030] Figure 4B This is a graph showing the temperature change of the cooling plate over time during abnormal handling.

[0031] Figure 5 It is to find and draw Figure 3A and Figure 3B The graph shows the cumulative area of ​​temperature under normal and abnormal data conditions.

[0032] Figure 6 It is to find and draw Figure 4A and Figure 4B The graph shows the cumulative area of ​​temperature under normal and abnormal data conditions.

[0033] Figure 7 This is a diagram illustrating the method for calculating the absolute value of the temperature difference per unit time in the change from temperature Ta to temperature Tz represented by temperature data.

[0034] Figure 8 It extracts and plots data per unit of time. Figure 4A , Figure 4B A graph showing the temperature of the data.

[0035] Figure 9 It is a graph plotting the cumulative absolute values ​​of the temperature difference over each elapsed time period.

[0036] Figure 10A It plots the score values ​​obtained from normal data for each data point.

[0037] Figure 10B It plots the score values ​​obtained from the outlier data for each data point. Detailed Implementation

[0038] Hereinafter, embodiments of the present invention will be described based on the accompanying drawings. Furthermore, in each of the following embodiments, the same or equivalent parts will be given the same reference numerals for description.

[0039] (First Embodiment)

[0040] The anomaly detection system of this embodiment will be described. Here, the example described is that the anomaly detection system is used as a system for anomaly detection in a coating / developing apparatus that performs a coating process of applying resist to a semiconductor wafer and a developing process.

[0041] After coating a semiconductor wafer with photoresist and then patterning it to obtain the desired pattern, the process involves sequentially performing a photoresist coating process, an exposure process, and a development process. The coating process involves applying photoresist onto the semiconductor wafer, stabilizing the photoresist by performing heating and cooling processes before and after coating. The subsequent exposure process involves exposing the wafer to light, using a photomask of a shape corresponding to the desired pattern to transfer the pattern. The development process is the process of forming the photoresist mask with the desired pattern through development. The process is reversed for negative and positive photoresists; however, with negative photoresists, development removes the unexposed areas, while with positive photoresists, development removes the photosensitive areas.

[0042] A coating / developing apparatus is used to perform the coating and developing processes. Since both the coating and developing processes involve heating and cooling, they are performed using the same coating / developing apparatus, but an exposure process occurs between the coating and developing processes. The exposure process is performed by removing the semiconductor wafer from the coating / developing apparatus and transporting it to the exposure apparatus.

[0043] The following describes the patterning of the resist, including the coating and developing processes using a coating / developing apparatus equipped with an anomaly detection system.

[0044] The coating process using a coating / developing apparatus, for example, is carried out through... Figures 1A-1I Perform as shown. First, as... Figure 1A In this way, the semiconductor wafer 1, before the resist coating is applied, is grasped as a workpiece by the conveyor arm 10 and moved to the mounting surface 21 of the hot plate 20, which is equivalent to a mounting plate, located in the heating device for the heating process. At this time, the heating device has been heated to the desired temperature for the heating process, such as 200°C, and the hot plate 20 is also exposed to this temperature, thus being heated to the temperature for the heating process.

[0045] The hot plate 20 includes a guide section 22 consisting of guide pins corresponding to the outer edge of the semiconductor wafer 1, and a lifting pin 23 for raising and lowering the semiconductor wafer 1. The guide section 22 is provided at six equal intervals along the outer edge of the semiconductor wafer 1. The lifting pin 23 is provided in a triangular shape at three central positions of the guide section 22 and is movable.

[0046] On the hot plate 20, such as Figure 1B Move the transport arm 10 as shown so that the center position of the semiconductor wafer 1 is aligned with the lifting pin 23. Then, release the grip of the transport arm 10 on the semiconductor wafer 1, as shown. Figure 1C After placing the semiconductor chip 1 on the lifting pin 23 as shown, as Figure 1DMove the conveyor arm 10 as shown so that only the semiconductor wafer 1 remains on the lifting pin 23. Then, as shown... Figure 1E As shown, the lifting pin 23 is lowered further, bringing the semiconductor wafer 1 into contact with the mounting surface 21 of the hot plate 20. Thus, the semiconductor wafer 1 is lowered while wafer misalignment is suppressed by the guide portion 22. Next, in this state, the semiconductor wafer 1 is placed in the heating apparatus for a desired time, thereby performing a heating process. Through this heating process, moisture adhering to the semiconductor wafer 1 can be dissipated.

[0047] Next, the semiconductor wafer 1, having completed the heating process, is grasped again by the conveyor arm 10 and moved into the cooling device for the cooling process. Then, it is moved onto the mounting surface 31 of the cooling plate 30, which is equivalent to a mounting plate, disposed within the cooling device. Then, through the aforementioned... Figures 1A-1E The same method, such as Figure 1F As shown, the guide portion 32 suppresses wafer misalignment, bringing the semiconductor wafer 1 into contact with the mounting surface 31 of the cooling plate 30. At this time, the cooling apparatus is cooled to the desired temperature for the cooling process, for example, 23°C, and the cooling plate 30 is exposed to this temperature, thus reaching the temperature required for the cooling process. Next, in this state, the semiconductor wafer 1 is placed in the cooling apparatus for a desired time, thereby performing the cooling process. Through this cooling process, the semiconductor wafer 1, heated by the aforementioned heating process, can be cooled.

[0048] Then, as Figure 1G As shown, a semiconductor wafer 1 is fed into a spin coater, and a photoresist 2 composed of a photosensitive resin is coated onto the surface of the semiconductor wafer 1.

[0049] Next, the semiconductor wafer 1 coated with resist 2 is again grasped as a workpiece by the conveyor arm 10 and conveyed into the heating device. Then, through the aforementioned... Figures 1A-1E The same method, such as Figure 1H As shown, a semiconductor wafer 1 coated with resist 2 is placed on a hot plate 20. At this time, the heating apparatus is heated to a desired temperature, such as 200°C, for the heating process, and the hot plate 20 is also heated to the temperature required for the heating process by being exposed to this temperature. Furthermore, the heating process is performed by placing the semiconductor wafer 1 in the heating apparatus for a desired time in this state. As a result, the semiconductor wafer 1 is heated, and the solvent contained in the resist 2 dissipates.

[0050] Then, the semiconductor wafer 1, after completing the heating process, is transported to the cooling device, and through... Figures 1A-1E The same method, such as Figure 1IAs shown, the semiconductor wafer 1 is brought into contact with the mounting surface 31 of the cooling plate 30. At this time, the cooling apparatus is cooled to the desired temperature for the cooling process, such as 23°C, and the cooling plate 30 is also cooled to the temperature for the cooling process by being exposed to this temperature. Furthermore, the cooling process is performed by placing the semiconductor wafer 1 in the cooling apparatus for a desired time in this state. Through this cooling process, the semiconductor wafer 1, which has been heated by the heating process described above, can be cooled.

[0051] Thus, the coating process is completed by the coating / developing apparatus. After the coating process is completed, the semiconductor wafer 1 is grasped by the transport arm 10 and transported to an exposure apparatus (not shown). Then, in the exposure apparatus, an exposure process is performed to expose the resist using a desired photomask.

[0052] Next, after the exposure process is completed, the semiconductor wafer 1 is again grasped by the transport arm 10 and transported to the coating / developing apparatus. Then, the coating / developing apparatus performs the development process. In the development process, firstly, after reducing the standing wave effect in the resist film generated during the exposure process through heating and cooling processes, the resist 2 is developed to leave a desired pattern. Then, the resist 2 is solidified through heating and cooling processes. Regarding this development process, the semiconductor wafer 1 is... Figures 1A-1I The coating process shown Figure 1H The process of replacing the coating of resist 2 with the development of resist 2 is carried out in the same way for both the heating and cooling processes. In this way, the patterning of resist 2 is completed.

[0053] Here, as described above, heating and cooling processes are performed in the coating and developing processes by the coating / developing apparatus. During these heating and cooling processes, the heating apparatus is set to a certain heating temperature and the cooling apparatus is set to a certain cooling temperature, but when the semiconductor wafer 1 is mounted, the hot plate 20 and the cooling plate 30 temporarily change from the heating temperature and the cooling temperature.

[0054] Specifically, during the heating process, when the semiconductor wafer 1 is placed on the hot plate 20, the temperature of the semiconductor wafer 1 is lower than that of the hot plate 20, so the hot plate 20 holding the semiconductor wafer 1 cools down due to heat conduction to the semiconductor wafer 1. Conversely, during the cooling process, when the semiconductor wafer 1 is placed on the cooling plate 30, the temperature of the semiconductor wafer 1 is higher than that of the cooling plate 30, so the cooling plate 30 holding the semiconductor wafer 1 heats up due to heat conduction from the semiconductor wafer 1. Therefore, the temperature of the hot plate 20, which has reached a certain heating temperature, or the cooling plate 30, which has reached a certain cooling temperature, temporarily changes.

[0055] Therefore, in the anomaly detection system of this embodiment, temperature data of the hot plate 20 and the cooling plate 30 are obtained, and anomaly detection is performed based on this temperature data. First, refer to Figure 2 The structure of the anomaly detection system is described.

[0056] like Figure 2 As shown, the anomaly detection system has temperature sensors 40 and 50 and a control unit 60.

[0057] Temperature sensor 40 is disposed on the back of hot plate 20 within coating / developing apparatus 100, and outputs a detection signal as temperature data corresponding to the temperature of hot plate 20. Temperature sensor 50 is disposed on the back of cooling plate 30 within coating / developing apparatus 100, and outputs a detection signal as temperature data corresponding to the temperature of cooling plate 30. The detection signals from these temperature sensors 40 and 50 are input to control unit 60.

[0058] The control unit 60 includes a microcomputer equipped with a CPU, ROM, RAM, I / O, etc., and performs anomaly detection based on the detection signals output from the temperature sensors 40 and 50. Specifically, the control unit 60 includes various functional units for anomaly detection, such as a waveform acquisition unit 61, an anomaly degree calculation unit 62, a threshold storage unit 63, and an anomaly determination unit 64. These various functional units will be described below, but before that, an overview of the anomaly detection performed by the anomaly detection system of this embodiment will be described while comparing it with conventional anomaly detection methods.

[0059] As described above, when the semiconductor wafer 1 is mounted on the hot plate 20 during the heating process or on the cooling plate 30 during the cooling process, the temperature of the hot plate 20 or the cooling plate 30 changes temporarily.

[0060] However, during abnormal handling such as when the semiconductor wafer 1 is transported while resting on the guide portions 22, 32 of the hot plate 20 or the cool plate 30, the temperature change is smaller compared to the normal handling where the semiconductor wafer 1 is properly positioned without resting on the guide portions 22, 32. That is, during abnormal handling, the contact area between the semiconductor wafer 1 and the hot plate 20 or the cool plate 30 is smaller than during normal handling, thus reducing heat conduction between the semiconductor wafer 1 and the hot plate 20 or the cool plate 30, and consequently reducing the temperature change of the hot plate 20 or the cool plate 30. The control unit 60 uses this phenomenon to perform abnormality detection.

[0061] For example, investigating the time-varying temperature of the hot plate 20 when the semiconductor wafer 1 is transported onto the hot plate 20 for a heating process, normal data during normal processing and abnormal data during abnormal processing become... Figure 3A , Figure 3BThe results shown are as described. Furthermore, the temperature change of the cooling plate 30 over time was investigated when the semiconductor wafer 1, after the heating process, was transported onto the cooling plate 30 for a cooling process; normal data during normal processing and abnormal data during abnormal processing were obtained. Figure 4A , Figure 4B The result shown. Additionally, in Figure 3A , Figure 3B , Figure 4A , Figure 4B The graph is simplified, but in reality, dozens of data points were obtained for both normal and abnormal data. Furthermore, an example is given where the heating temperature is set to 200℃ and the cooling temperature to 23℃, but the heating and cooling temperatures are arbitrary.

[0062] like Figure 3A , Figure 3B As shown, during both normal and abnormal processing, the hot plate 20, initially heated to 200°C, exhibits a temperature change as the semiconductor wafer 1 is placed, followed by a decrease in temperature, then a rise, before stabilizing at 200°C. However, according to... Figure 3A , Figure 3B It can be seen that the temperature change is larger during normal processing, while the temperature change is smaller during abnormal processing compared to normal.

[0063] On the other hand, such as Figure 4A , Figure 4B As shown, during both normal and abnormal processing, the cooling plate 30, with a cooling temperature of 23°C, exhibits a temperature change as the semiconductor wafer 1 is placed, rising, then falling, and stabilizing at 23°C again. However, according to... Figure 4A , Figure 4B It can be seen that the temperature change is larger during normal processing, while the temperature change is smaller during abnormal processing compared to normal.

[0064] In the anomaly detection method described in Patent Document 1, which is prior art, the cumulative area of ​​the temperature change curve and the range enclosed by the set temperature (here, 23°C) is calculated, and anomaly detection is performed by comparing it with a threshold.

[0065] Therefore, the results were calculated and plotted. Figure 3A Normal data shown under normal processing conditions and Figure 3B The accumulated area of ​​temperature under each of the abnormal data conditions shown in the anomaly handling examples becomes... Figure 5The results are shown in the figure. As the figure illustrates, the cumulative temperature area of ​​normal data is concentrated around 9965–9975, while the cumulative temperature area of ​​abnormal data is concentrated around 9950–9960. In this case, the threshold can be set between the values ​​of the concentrated temperature areas of normal and abnormal data. Therefore, it is possible to reliably distinguish between wafers that have undergone normal processing and wafers that have undergone abnormal processing.

[0066] However, the results were calculated and plotted. Figure 4A Normal data shown under normal processing conditions and Figure 4B The accumulated temperature area under each of the abnormal data conditions shown in the abnormal handling case becomes... Figure 6 The results are shown in the figure. As the figure illustrates, the accumulated temperature area for normal data is concentrated around 1380 or 1385, while the accumulated temperature area for abnormal data is concentrated around 1381, in between. Thus, the accumulated temperature area for abnormal data is concentrated between the two values ​​where the accumulated temperature area for normal data is concentrated, resulting in no difference between the accumulated temperature areas of normal and abnormal data. Therefore, it is impossible to set a threshold for distinguishing between normal and abnormal data based on these results, and it is no longer possible to reliably distinguish between normally processed and abnormally processed wafers. Consequently, situations arise where wafers that have been processed normally are incorrectly detected as abnormally processed wafers, requiring manufacturing interruptions despite normal processing, thus hindering production.

[0067] Therefore, in this embodiment, each temperature data point is divided into units of time, and the absolute value of the temperature difference in each unit of time is calculated. Furthermore, based on the cumulative value of the absolute value of this temperature difference, a score representing the degree of anomaly is calculated based on Hotelling's theory. By using this score, it is determined whether the wafer has been processed normally or abnormally, thereby performing anomaly detection.

[0068] Specifically, such as Figure 7 As shown, for each specified unit time in the temperature data, from the initial temperature Ta at the start of the temperature change to the stable temperature Tz after the temperature change ends, the absolute value of the temperature difference is calculated, and its cumulative value I is calculated as follows. Furthermore, T(n) refers to any time during the period from the start of the temperature change to its end, and T(n+1) refers to the time after one unit time has elapsed from T(n).

[0069] [Mathematical Expression 1]

[0070]

[0071] Extract and plot the above data per unit time. Figure 4A, Figure 4B The temperatures of the data shown are as follows: Figure 8 That's how it's represented. Furthermore, calculate the temperature difference between adjacent plots of each data point, take this as the absolute value of the temperature difference, and plot the cumulative value of this temperature difference over time. Then, as shown... Figure 9 That's how it's expressed.

[0072] Here, the cumulative value of the absolute value of the temperature difference per unit time is calculated from the initial temperature Ta at the start of the temperature change to the final temperature Tz after the temperature change ends and stabilizes. Figure 9 The rightmost plotted value is used to calculate the score. Various methods can be used to score normal and abnormal data, but here we use Hotelling's theory to calculate the score based on cumulative values ​​for anomaly detection.

[0073] This score serves as an indicator of the degree of abnormality in the heating or cooling process. A graph is plotted for each data point, based on the scores obtained from dozens of normal and abnormal data points respectively. Figure 10A , Figure 10B .like Figure 10A As shown, the scores for normal data are all low, while the scores for abnormal data are all high compared to the scores for normal data. Therefore, a threshold is set between the scores for normal data and the scores for abnormal data. If the score is lower than the threshold, it can be identified as normal data; if the score is higher than the threshold, it can be identified as abnormal data. In the anomaly detection system of this embodiment, anomaly detection is performed using this method. Hereinafter, based on this anomaly detection method, the various functional units included in the control unit 60, namely the waveform acquisition unit 61, the anomaly calculation unit 62, the threshold storage unit 63, and the anomaly determination unit 64, will be described in detail.

[0074] The waveform acquisition unit 61 inputs the detection signal output by the temperature sensor 40 or the temperature sensor 50. The detection signal input from the temperature sensor 40 or the temperature sensor 50 becomes raw waveform data representing the temperature of the hot plate 20 or the cooling plate 30.

[0075] The anomaly calculation unit 62 calculates the anomaly degree in the heating or cooling process based on the original waveform data of the temperature of the hot plate 20 or the cooling plate 30 obtained by the waveform acquisition unit 61. The anomaly degree referred to here is the score value mentioned above. Based on the aforementioned mathematical formula 1, according to the original waveform data, the absolute value of the temperature difference is calculated for each predetermined unit of time from the temperature Ta at the beginning of the temperature change to the temperature Tz after the temperature change ends and stabilizes, and its cumulative value is calculated. The score value is then obtained by calculating the negative logarithm.

[0076] The threshold storage unit 63 is a portion that stores a threshold set between the score values ​​of normal data and the score values ​​of abnormal data, and is composed of a non-movable physical recording medium such as a memory. The threshold is determined in advance through experiments, etc. A normal model based on normally processed data is prepared, and scores are calculated for each data point using both normally processed and abnormally processed data. Furthermore, the threshold is set as the value between the score value obtained from normal data after normal processing and the score value obtained from multiple abnormal data after abnormal processing.

[0077] That is, by collecting multiple score values ​​obtained from normal data under normal processing, it is possible to create a distribution of score values ​​under normal processing, i.e., a normal distribution. If a score value deviates from this distribution, it can be considered a score value under abnormal processing. Therefore, the boundary value between the score value obtained from normal data under normal processing and the score value obtained from multiple abnormal data under abnormal processing, such as the maximum value of the score value contained in the normal distribution, is set as a threshold.

[0078] In addition, the threshold is a value set separately for each heating process and each cooling process of the coating process and the developing process, and the separately set threshold is stored in the threshold storage unit 63.

[0079] The anomaly determination unit 64 determines whether the score value is a score value for normal data or a score value for abnormal data by comparing the anomaly degree, i.e., the score value, calculated by the anomaly degree calculation unit 62 with the corresponding threshold value stored in the threshold storage unit 63. In other words, the anomaly determination unit 64 determines whether the data waveform acquired by the waveform acquisition unit 61 is normal or abnormal by comparing the anomaly degree with the threshold value.

[0080] For example, if the score value is calculated based on the original waveform data of the temperature of the cooling plate 30 in the cooling process of the coating process, it is compared with a threshold set for the cooling process of the coating process. If the score value is below the threshold, it is determined to be a normal score value; if it is above the threshold, it is determined to be an abnormal score value. Furthermore, if the abnormality determination unit 64 determines that the score value is abnormal, it transmits it externally. For example, an abnormality is reported via a reporting device (not shown). The reporting device can be a device that performs various reports, such as visual reports (e.g., display on a screen) or auditory reports (e.g., using a buzzer).

[0081] As explained above, in the anomaly detection system of this embodiment, during the heating and cooling processes, the cumulative value of the absolute value of the temperature difference per unit time is calculated based on the temperature data of the hot plate 20 and the cooling plate 30, thereby obtaining a score representing the degree of anomaly. Furthermore, by comparing the score with a preset threshold, it is determined whether the score represents normal data or abnormal data. In this way, by using the score obtained from the cumulative value of the absolute value of the temperature difference per unit time as the degree of anomaly detection, accurate anomaly detection can be performed. Moreover, according to this anomaly detection method, anomaly detection can be performed whenever a heating or cooling process is performed, thus enabling faster anomaly detection.

[0082] (Other implementation methods)

[0083] The present invention has been described according to the above embodiments, but is not limited to these embodiments, and includes various modifications and equivalent variations. In addition, various combinations and forms, and further, other combinations and forms that include only one element, or include more or less of it, also fall within the scope and spirit of the present invention.

[0084] In the above embodiments, the heating and cooling processes during the coating and development of resist 2 on the semiconductor wafer 1 are used as examples, and the carrier plates such as the hot plate 20 and the cooling plate 30 used to mount the semiconductor wafer 1 are used as examples for temperature measurement. However, these are just examples, and the present invention can also be applied when determining whether the waveform data is normal or abnormal based on waveform data representing the temperature measurement result of other objects.

[0085] In addition, anomaly detection can also be performed whenever time-varying waveform data is used for anomaly detection. That is, the anomaly score is calculated by accumulating the absolute values ​​of the differences in data values ​​per unit time of the waveform data, and this score is compared to a pre-set threshold to perform anomaly detection. As an example of using waveform data for anomaly detection, detecting whether a gas flow rate is normal or abnormal can be given. In this case, the waveform data of the gas flow rate shows the change in flow rate relative to a certain value, so a score is obtained by accumulating the absolute values ​​of the differences in gas flow rate per unit time of the waveform data. Because anomaly detection can be performed every time, it can be done more quickly.

[0086] The control unit and method described in this invention can also be implemented by a dedicated computer consisting of a processor and memory programmed to perform one or more functions, embodied in a computer program. Alternatively, the control unit and method described in this invention can also be implemented by a dedicated computer consisting of a processor configured with one or more dedicated hardware logic circuits. Alternatively, the control unit and method described in this invention can also be implemented by one or more dedicated computers consisting of a combination of a processor and memory programmed to perform one or more functions and a processor configured with one or more hardware logic circuits. Furthermore, the computer program can also be stored as instructions executable by the computer in a computer-readable non-removable tangible recording medium.

Claims

1. An anomaly detection system, characterized in that, have: The waveform acquisition unit (61) acquires waveform data that changes in a time sequence; The anomaly calculation unit (62) calculates the anomaly degree based on the cumulative value obtained by accumulating the absolute value of the difference between the data values ​​per unit time based on the waveform data obtained by the waveform acquisition unit. as well as The anomaly determination unit (64) determines whether the waveform data is normal or abnormal based on the anomaly degree calculated by the anomaly degree calculation unit. The waveform data that changes over time above represents the temperature measurement results of the object being measured. The aforementioned anomaly calculation unit, let T(n) be any time point between the start time and the end of the change, and let T(n+1) be the time point after the aforementioned unit of time from that arbitrary time point. Through... [Mathematical Expression 1] Calculate the cumulative value I of the absolute value of the temperature difference per unit time from the temperature Ta at the beginning of the temperature change to the temperature Tz at the end of the temperature change, as represented by the waveform data above.

2. The anomaly detection system as described in claim 1, characterized in that, It has a threshold storage unit (63) that stores a threshold for comparison with the above-mentioned anomaly degree; The above-mentioned anomaly determination unit determines whether the above-mentioned waveform data is normal or abnormal by comparing the above-mentioned anomaly degree with the above-mentioned threshold.

3. The anomaly detection system as described in claim 2, characterized in that, The above-mentioned anomaly calculation section calculates the anomaly degree when the waveform data is normal and the anomaly degree when the waveform data is abnormal, respectively. The threshold stored in the threshold storage section is a value that is preset as the value between the anomaly degree when the waveform data is normal and the anomaly degree when the waveform data is abnormal.

4. The anomaly detection system as described in claim 1, characterized in that, The temperature measurement object is the mounting plate (20, 30) on which the semiconductor wafer (1) is placed during heating or cooling.

Images