End point detection during mixing

Abstract

In some implementations, the device can identify a pseudo-steady-state endpoint, indicating the end of a pseudo-steady-state associated with the mixing process, based on spectral data. The device can identify a reference block and a test block from the spectral data based on the pseudo-steady-state endpoint. The device can generate raw detection signals associated with the reference block and raw detection signals associated with the test block. The device can generate a statistical detection signal based on the raw detection signals associated with the reference block and the test block. The device can determine whether the mixing process has reached a steady state based on the statistical detection signal.

Description

Technical Field

[0001] This application relates to a mixing process; more specifically, to the detection of the endpoint of a mixing process. Background Technology

[0002] A mixing process (e.g., a mixing process associated with the manufacture of a pharmaceutical product) may involve one or more state transitions, such as a transition from an unstable state (e.g., a non-homogeneous state of mixing where the properties of the mixture change over time) to a stable state (e.g., a homogeneous state of mixing where the properties of the mixture remain substantially constant over time). For example, a mixing process may involve a transition of the spectral properties of the mixture from an unstable state (e.g., at the start of the mixing process) to a stable state (e.g., indicating the completion of the mixing process). Summary of the Invention

[0003] Some embodiments described herein relate to a method. The method may include receiving spectral data associated with a mixing process by a device. The method may include identifying a pseudo-steady-state endpoint based on the spectral data, the pseudo-steady-state endpoint indicating the end of a pseudo-steady-state associated with the mixing process. The method may include identifying a reference block and a test block from the spectral data based on the pseudo-steady-state endpoint. The method may include generating a raw detection signal associated with the reference block and a raw detection signal associated with the test block by the device. The method may include generating a statistical detection signal based on the raw detection signal associated with the reference block and the raw detection signal associated with the test block by the device. The method may include determining, based on the statistical detection signal, whether the mixing process has reached a steady state.

[0004] Some embodiments described herein relate to an apparatus. The apparatus may include one or more memories and one or more processors coupled to the memories. The apparatus may be configured to receive spectral data associated with a mixing process. The apparatus may be configured to identify a pseudo-steady-state endpoint based on the spectral data, the pseudo-steady-state endpoint indicating the end of a pseudo-steady-state associated with the mixing process. The apparatus may be configured to identify a reference block and a test block from the spectral data based on the pseudo-steady-state endpoint. The apparatus may be configured to generate a raw detection signal associated with the reference block and a raw detection signal associated with the test block. The apparatus may be configured to generate a statistical detection signal based on the raw detection signal associated with the reference block and the raw detection signal associated with the test block. The apparatus may be configured to determine whether the mixing process has reached a steady state based on the statistical detection signal.

[0005] Some embodiments described herein relate to a non-transitory computer-readable medium storing an instruction set for a device. When run by one or more processors of the device, the instruction set enables the device to receive spectral data associated with a mixing process. When run by one or more processors of the device, the instruction set enables the device to identify a pseudo-steady-state endpoint based on the spectral data, the pseudo-steady-state endpoint indicating the endpoint of a pseudo-steady-state associated with the mixing process. When run by one or more processors of the device, the instruction set enables the device to identify a reference block and a test block from the spectral data based on the pseudo-steady-state endpoint. When run by one or more processors of the device, the instruction set enables the device to generate a raw detection signal associated with the reference block and a raw detection signal associated with the test block. When run by one or more processors of the device, the instruction set enables the device to generate a statistical detection signal based on the raw detection signals associated with the reference block and the test block. When run by one or more processors of the device, the instruction set enables the device to determine whether the mixing process has reached a steady state based on the statistical detection signal. Attached Figure Description

[0006] Figures 1A-1H This is a diagram associated with an example implementation of hybrid process endpoint detection as described herein.

[0007] Figure 2 This is a diagram of an example environment in which the systems and / or methods described herein can be implemented.

[0008] Figure 3 yes Figure 2 A diagram of example components of one or more devices.

[0009] Figure 4 This is a flowchart of an example process related to endpoint detection in a hybrid process, as described in this article. Detailed Implementation

[0010] The following detailed description of exemplary embodiments is taken with reference to the accompanying drawings. The same reference numerals in different drawings may identify the same or similar elements. The following description uses a spectrometer as an example. However, the techniques, principles, procedures, and methods described herein can be used with any sensor, including but not limited to other optical and spectral sensors.

[0011] The homogeneity (sometimes called uniformity) of compounds produced by mixing processes (e.g., those related to the manufacture of pharmaceutical products) should be monitored to ensure the quality and performance of the mixing process. Achieving acceptable homogeneity allows the compound to be effectively utilized, for example, in subsequent process steps or made available to the consumer. Conversely, poor homogeneity can lead to unusable or rejected compounds, meaning that resources dedicated to the mixing process are wasted.

[0012] As described above, a mixing process can include a transition of the spectral properties of the mixture from an unstable state (e.g., a non-uniform state in which the properties of the mixture change over time) to a stable state (e.g., a homogeneous state in which the properties of the mixture remain substantially constant over time), the stable state indicating that mixing has been achieved. Therefore, accurate and reliable detection of the stable state based on the spectral characteristics of the mixture can improve both the performance of the mixing process (e.g., by ensuring adequate mixing) and the efficiency of the mixing process (e.g., by enabling the mixing process to terminate immediately after mixing is achieved).

[0013] A conventional technique for detecting the endpoint of a mixing process based on spectral characteristics is the use of a moving F-test. According to this technique, the difference between two near-infrared (NIR) spectral blocks collected during the mixing process is evaluated to determine when a steady state is reached, and thus when the endpoint of the mixing process is reached. However, this conventional technique does not provide robust endpoint detection and often detects the endpoint prematurely (i.e., before the mixing process actually reaches a steady state). Other conventional techniques for detecting the endpoint of a mixing process can use another type of moving block analysis, such as the standard deviation of the moving block, the relative standard deviation of the moving block, or the mean of the moving block. However, these other types of moving block analyses rely on historical spectral data to set the threshold for detecting the steady state. Furthermore, these other types of moving block analyses may also detect the endpoint prematurely (e.g., similar to the moving F-test). Due to these problems, conventional techniques can be unreliable (e.g., due to early endpoint detection) and / or undesirably complex (e.g., due to reliance on historical data).

[0014] Some embodiments described herein include techniques and apparatus for providing improved detection of the endpoint of a mixing process. In some embodiments, the apparatus may receive spectral data associated with the mixing process. The apparatus may identify pseudo-steady-state endpoints based on the spectral data, and may identify reference blocks and test blocks from the spectral data based on pseudo-steady-state endpoints. The apparatus may generate raw detection signals associated with the reference block and raw detection signals associated with the test block, and then generate statistical detection signals based on the raw detection signals. The apparatus may then determine whether the mixing process has reached a steady state based on the statistical detection signals. In some embodiments, the techniques and apparatus described herein provide robust endpoint detection based on spectral data collected during the mixing process (e.g., independent of historical spectral data). Further details are provided below.

[0015] Figures 1A-1H This is a graph associated with the detection of the endpoint of the mixing process, as described in this article. Figure 1A and Figure 1B This is a diagram illustrating an example implementation 100 for endpoint detection in a mixing process. (See diagram for example.) Figure 1A and Figure 1B As shown, Example Implementation 100 includes a spectrometer 210, a detection device 220, and a user device 230.

[0016] like Figure 1A As indicated by reference numeral 102 in the figures, detection device 220 can receive spectral data associated with the mixing process. For example, as shown, spectrometer 210 can measure spectral data at a given time during the execution of the mixing process and can provide the spectral data to detection device 220. In some embodiments, the spectral data includes spectra (e.g., multivariate time-series data, such as NIR spectra) measured by spectrometer 210 during the execution of the mixing process.

[0017] In some embodiments, the detection device 220 can receive spectral data in real-time or near real-time during the mixing process. For example, the detection device 220 can receive spectral data measured by the spectrometer 210 during the mixing process in real-time or near real-time relative to the spectrometer 210. In some embodiments, the detection device 220 can determine whether the end point of the mixing process has been reached based on the spectral data, as described herein.

[0018] In some embodiments, as indicated by reference numeral 104, the detection device 220 can identify the pseudo-steady-state endpoint based on spectral data. A pseudo-steady-state is the state of a mixed process between an unstable and a stable state. In other words, a pseudo-steady-state is a transitional (or meta-)state between an unstable and a stable state. In a pseudo-steady-state, the mixed process is neither in an unstable nor a stable state. The pseudo-steady-state endpoint is the point in time indicating the end of the pseudo-steady-state. In some embodiments, the detection device 220 identifies the pseudo-steady-state endpoint so that, for the purpose of identifying the endpoint of the mixed process, the spectral data corresponding to the unstable state of the mixed process can be ignored, thereby reducing or eliminating noise generated by the spectral data corresponding to the unstable state of the mixed process, and thus improving the reliability of the mixed process endpoint detection.

[0019] In some implementations, detection device 220 may identify junction points based on the endpoint of a pseudo-steady state. A junction point is the point in time when the mixing process transitions from an unstable state to a pseudo-steady state. That is, a junction point is the point in time when the pseudo-steady state begins (e.g., the point in time when the mixing process enters a pseudo-steady state). In some implementations, detection device 220 may identify junction points to enable visualization of the evolution of the mixing process. For example, detection device 220 may utilize junction points to provide visualization of the mixing process, allowing a user (e.g., a user of user device 230) to view the evolution history of the mixing process (e.g., for real-time or near-real-time monitoring of the mixing process, for performing post-run diagnostics of the mixing process, for setting parameters associated with the mixing process, etc.).

[0020] In some implementations, in association with identifying pseudo-steady-state endpoints (and junctions), detection device 220 can perform a rolling F-test of the two blocks of spectral data. For example, in some implementations, detection device 220 can identify a pseudo-steady-state endpoint by calculating a first time point in the mixing process where a p-value (e.g., the probability of obtaining an outcome at least as extreme as the observed outcome, assuming the null hypothesis is correct) satisfies (e.g., greater than or equal to) a p-value threshold. Therefore, in some implementations, detection device 220 calculates the p-value between the moving two blocks of spectral data such that detection device 220 calculates the p-value at time steps in the mixing process. Detection device 220 then identifies the pseudo-steady-state endpoint as the time point (e.g., the first time point) where the p-value satisfies the p-value threshold. In some implementations, the p-value threshold can be user-selectable or configurable (e.g., to ensure desired robustness). The p-value threshold can be, for example, 0.01, 0.02, 0.05, or another suitable value.

[0021] In some implementations, the detection device 220 may identify junctions based on spectral data and the pseudo-steady-state endpoint. For example, in some implementations, when using p-values, the detection device 220 may identify junctions by (1) identifying the time point in time steps where the minimum p-value occurs (e.g., the time step range between the first time step of the mixing process and the pseudo-steady-state endpoint) and (2) subtracting the block size (e.g., the size of the moving double block in time steps) from the identified time point.

[0022] Figure 1C This is a graph illustrating an example of how p-values ​​calculated based on a two-block rolling F-test using spectral data are associated with identifying pseudo-steady-state endpoints and junctions. Figure 1C In this context, the p-value (displayed on the y-axis as a logarithm to base 10) is calculated by the detection device 220 at each time step of the mixing process (shown on the x-axis as the time point at the end of the moving biplane). For example... Figure 1C As shown, the moving pairs are initially significantly different from each other, resulting in small p-values. However, as the mixing process continues, the differences between the moving pairs decrease, meaning the p-values ​​increase over time. Figure 1C In the example shown, the first time step where the p-value is greater than the p-value threshold is time step 43. Therefore, in this example, the detection device 220 identifies the pseudo-steady-state endpoint as time step 43. It is worth noting that the conventional techniques described above for endpoint detection might identify time step 43 as the point in time when the mixing process is in a steady state. Furthermore, in Figure 1C In this case, the minimum p-value is at time step 30, and the block size is 15, meaning the junction is identified at time step 15. In some implementations, the junction is determined based on the minimum p-value because when the mixing process transitions from an unstable state to a pseudo-stable state, the minimum p-value is at the junction between the moving two blocks, one of which is in an unstable state while the other is in a pseudo-stable state; therefore, the moving two blocks should have the minimum p-value.

[0023] In some implementations, detection device 220 may use F-values ​​and F-value thresholds to identify pseudo-steady-state endpoints and (optionally) junction points (e.g., in addition to using p-values, or not using p-values). For example, in some implementations, detection device 220 may identify pseudo-steady-state endpoints by calculating a first time point in the mixing process where the F-value (e.g., a value calculated using regression analysis to determine whether the means between two populations are significantly different) satisfies (e.g., is less than) the F-value threshold. Therefore, in some implementations, detection device 220 calculates the F-value between the moving biblocks such that detection device 220 calculates the F-value at time steps in the mixing process and identifies the pseudo-steady-state endpoint as the time point where the F-value satisfies the F-value threshold. In some implementations, the F-value thresholds associated with identifying pseudo-steady-state endpoints may be objectively generated (e.g., rather than user-selectable or configurable).

[0024] In some implementations, when using the F-value, the detection device 220 can identify the junction point by (1) identifying the time point in time steps where the maximum F-value occurs (e.g., the time step range between the first time step of the mixing process and the end of the pseudo-steady state), and (2) subtracting the block size (e.g., the size of the moving double block in time steps) from the identified time point. In this example, the detection device 220 identifies the junction point by (1) identifying the time point in time steps where the maximum F-value occurs (e.g., the time step range between the first time step of the mixing process and the end of the pseudo-steady state), and (2) subtracting the block size from the identified time point.

[0025] Figure 1D This is a graph illustrating an example of how a rolling F-test based on two blocks of spectral data using F-values ​​can be used to identify pseudo-steady-state endpoints. Figure 1D In this context, the F-value (shown on the y-axis) is calculated by the detection device 220 at each time step of the mixing process (shown on the x-axis as the number of revolutions associated with the mixing process). For example... Figure 1D As shown, the moving pairs of blocks are initially significantly different from each other, resulting in large F-values. However, as the mixing process continues, the differences between the moving pairs decrease, meaning the F-values ​​decrease over time. Figure 1D In the example shown, the F value is less than the F value threshold (e.g., Figure 1D In step 2.5), the first time step is time step 42. Therefore, in this example, detection device 220 identifies the pseudo-steady state endpoint as time step 42. Figure 1D In this case, the maximum F value is at time step 30, and the block size is 15, which means that the junction is identified at time step 15.

[0026] In some implementations, the detection device 220 can be configured to use p-values ​​and F-values ​​in combination to identify pseudo-steady-state endpoints and junctions. In this case, the detection device 220 can use p-values ​​to identify the first pseudo-steady-state endpoint and the first junction, and can use F-values ​​to identify the second pseudo-steady-state endpoint and the second junction. Here, the detection device 220 can identify the pseudo-steady-state endpoint for further data processing as the latest time point (e.g., the most conservative) among the first and second pseudo-steady-state endpoints, and similarly, can identify the junction for further data processing as the latest time point among the first and second junctions. This improves the reliability and robustness of endpoint detection.

[0027] In some implementations, the detection device 220 can perform a rolling F-test on the entire spectrum of the spectral data within the moving biblock, associated with identifying the endpoint of a pseudo-steady state. For example, the detection device 220 can perform automatic scaling or another type of preprocessing on the spectral data within the moving biblock, and the rolling F-test can be performed using the entire preprocessed spectrum of the spectral data within the moving biblock. It is noteworthy that performing a rolling F-test across the entire spectrum can, in some cases, reduce the computational complexity performed by the detection device 220.

[0028] Additionally or alternatively, detection device 220 may use a principal component analysis (PCA) model generated based on spectral data to perform a rolling F-test. For example, in some embodiments, detection device 220 may perform PCA dimensionality reduction on the spectral data within the moving biblock to identify a set of principal components (e.g., two principal components, three principal components, four principal components, etc.) associated with the spectral data within the moving biblock. Detection device 220 may then (dynamically) generate a PCA model based on the identified principal components, and may use the PCA model to calculate PCA scores associated with the moving biblock. Detection device 220 may then perform a rolling F-test based on the PCA scores calculated using the PCA model.

[0029] Additionally or alternatively, detection device 220 may use a PCA model generated based on historical spectral data (e.g., spectral data associated with iterations of previously performed mixing processes) to perform a rolling F-test. For example, in some embodiments, detection device 220 may perform PCA dimensionality reduction on the historical spectral data to identify a set of principal components (e.g., two principal components, three principal components, four principal components, etc.) associated with the historical spectral data. Detection device 220 may then generate a historical PCA model based on the identified principal components. Here, detection device 220 may use the historical model to calculate PCA scores associated with biblocks of movement of the spectral data. Detection device 220 may then perform a rolling F-test based on the PCA scores calculated using the historical PCA model. It is worth noting that the use of historical spectral data or a historical PCA model is not mandatory, but it can be used to improve the reliability and robustness of endpoint detection. In some embodiments, the detection device 220 can perform PCA dimensionality reduction on spectral data (i.e., spectral data associated with the current iteration of the mixing process) or historical spectral data, and then perform an F-test between the test block and the reference block of the spectral data using specific principal components (e.g., PC1, PC2, etc.). In some embodiments, specific principal components can be used to perform a moving F-test. In this way, the analysis can be more chemically characterized because the specific principal components will correspond to specific chemical information.

[0030] Back Figure 1A As indicated by reference numeral 106 in the accompanying drawings, detection device 220 can identify reference blocks and test blocks from spectral data based on a pseudo-steady-state endpoint. A reference block (also called a reference unit) is a block of spectral data starting from or near the pseudo-steady-state endpoint. For example, a reference block may include a spectrum of a specific number of time steps (e.g., 30 time steps) or a block size (e.g., two block sizes) starting from a time point corresponding to the pseudo-steady-state endpoint. In some embodiments, the reference block remains fixed (e.g., does not move) when detection device 220 receives additional spectral data during the mixing process.

[0031] A test block is a block of spectral data that can be compared with a reference block associated with determining whether the mixing process has reached a steady state. In some implementations, the test block (also called a test unit) moves dynamically based on newly received spectral data. For example, a first test block may include a spectrum of a specific number of time steps (e.g., 15 time steps) or a block size (e.g., one block size) starting from a time point corresponding to the end of the pseudo-steady state. In this example, a second test block (e.g., another iteration for generating the original detection signal associated with the test block) may be updated to include a spectrum of a specific number of time steps (e.g., 15 time steps) or a block size (e.g., one block size) starting from a time point corresponding to a time point after the end of the pseudo-steady state (e.g., the end of the first test block, the midpoint of the first test block, a time point one or more steps later than the end of the pseudo-steady state, etc.). In this way, the test block can move dynamically over time (e.g., the reference block may remain stationary compared to the reference block). In some implementations, the static reference block and the dynamic test block form a so-called pseudo-moving block (PMB) architecture.

[0032] As shown by reference numeral 108 in the accompanying drawings, the detection device 220 can generate a raw detection signal associated with a reference block and a raw detection signal associated with a test block. The raw detection signal associated with the reference block and the raw detection signal associated with the test block are signals that the detection device 220 generates based on statistical detection signals used to determine whether the mixing process has reached its endpoint (e.g., whether the mixing process is in a steady state).

[0033] In some embodiments, the detection device 220 generates the original detection signal (e.g., the original detection signal associated with the reference block and / or the original detection signal associated with the test block) by calculating the travel space volume (TSV) associated with a block of spectral data (e.g., a reference block or a test block). The TSV is a hyperdimensional volume scanned by multivariate time-series spectral data within a time frame corresponding to the block, where the block size is defined by the number of time steps covered by the block. Therefore, the TSV of the reference block can be a hyperdimensional volume scanned by spectral data corresponding to the reference block (e.g., the spectrum over 30 time steps from the end of the pseudo-steady state), and the TSV of the test block can be a hyperdimensional volume scanned by spectral data corresponding to the test block (e.g., the spectrum over 15 time steps of the moving test block). In some embodiments, the detection device 220 can calculate the TSV by calculating the volume of the convex hull of the scan space within the block. Alternatively or additionally, the detection device 220 may calculate the TSV by calculating the volume of a hyperdimensional ellipsoid (e.g., a multivariate confidence ellipsoid based on a Hotelling T2 distribution). Alternatively or additionally, the detection device 220 may calculate the TSV by calculating the product of the standard deviations of each principal component of the spectral data corresponding to the block (e.g., when the detection device 220 is configured to perform PCA dimensionality reduction). Alternatively or additionally, the detection device 220 may calculate the TSV by calculating the product of the standard deviations of each wavelength of the spectral data corresponding to the block (e.g., when the detection device 220 is configured to use the entire spectrum of the spectral data).

[0034] Additionally or alternatively, the detection device 220 generates the original detection signal (e.g., the original detection signal associated with the reference block and / or the original detection signal associated with the test block) by calculating the single-block standard deviation (IBSD) associated with a block of spectral data (e.g., a reference block or a test block). Thus, the IBSD of the reference block can be the IBSD associated with the spectral data corresponding to the reference block, and the IBSD of the test block can be the IBSD associated with the spectral data corresponding to the test block. In some embodiments, the detection device 220 can calculate the IBSD by calculating the sum of squares of the standard deviations of each principal component of the spectral data corresponding to the block (e.g., when the detection device 220 is configured to perform PCA dimensionality reduction). Additionally or alternatively, the detection device 220 can calculate the IBSD by calculating the sum of squares of the standard deviations of each wavelength of the spectral data corresponding to the block (e.g., when the detection device 220 is configured to use the entire spectrum of the spectral data).

[0035] like Figure 1BAs shown by reference numeral 110 in the accompanying drawings, the detection device 220 can generate a statistical detection signal based on the original detection signal associated with the reference block and the original detection signal associated with the test block. The statistical detection signal is a signal that the detection device 220 can use to determine whether the mixing process has reached a steady state.

[0036] In some implementations, in association with generating a statistical detection signal, the detection device 220 can perform a rolling F-test between the reference block and the test block using both the original detection signal associated with the reference block and the original detection signal associated with the test block. For example, in some implementations, the detection device 220 calculates a p-value between the reference block and the test block such that the detection device 220 calculates the p-value throughout the mixing process. Here, the statistical detection signal is represented by the p-value calculated by the detection device 220 throughout the mixing process. As another example, in some implementations, the detection device 220 calculates an F-value between the reference block and the test block such that the detection device 220 calculates the F-value throughout the mixing process. Here, the statistical detection signal is represented by the F-value calculated by the detection device 220 throughout the mixing process. In some implementations, the statistical detection signal (e.g., instead of the original detection signal) is used to improve the robustness of the endpoint detection process.

[0037] As shown by reference numeral 112 in the attached figure, the detection device 220 can determine whether the mixing process has reached a steady state based on a statistical detection signal. For example, the detection device 220 can determine whether the statistical detection signal (e.g., p-value, F-value, etc.) satisfies a condition associated with a steady-state threshold. As a specific example, if the statistical detection signal is a p-value signal, the detection device 220 can determine whether the value of the statistical detection signal is less than or equal to the p-value steady-state threshold. Here, if the detection device 220 determines that the value of the statistical detection signal is greater than the p-value steady-state threshold, the detection device 220 can determine that the mixing process has not reached a steady state. Conversely, if the detection device 220 determines that the value of the statistical detection signal is less than or equal to the p-value steady-state threshold, the detection device 220 can determine that the mixing process has reached a steady state. As another example, if the statistical detection signal is an F-value signal, the detection device 220 can determine whether the value of the statistical detection signal is greater than or equal to the F-value steady-state threshold. Here, if the detection device 220 determines that the value of the statistical detection signal is less than the F-value steady-state threshold, the detection device 220 can determine that the mixing process has not reached a steady state. Conversely, if the detection device 220 determines that the value of the statistical detection signal is greater than or equal to the F-value steady-state threshold, then the detection device 220 can determine that the mixing process has reached a steady state.

[0038] In some implementations, the condition associated with the steady-state threshold may require a specific number of consecutive values ​​of the statistical detection signal to satisfy the steady-state threshold (e.g., to improve the robustness and reliability of endpoint detection). For example, the condition may require seven consecutive values ​​of the statistical detection signal to satisfy the steady-state threshold (e.g., seven consecutive p-values ​​less than or equal to the p-value steady-state threshold, seven consecutive F-values ​​greater than or equal to the p-value steady-state threshold, six consecutive p-values ​​less than or equal to the p-value steady-state threshold, six consecutive F-values ​​greater than or equal to the p-value steady-state threshold, etc.). In some implementations, the required number of consecutive values ​​may be selected based on a binomial probability function (e.g., to sufficiently reduce the likelihood of false positive endpoint detection). In some implementations, the steady-state threshold may be user-selectable or configurable (e.g., when the statistical detection signal is a p-value signal). Alternatively, in some implementations, the steady-state threshold may be objectively calculated by the detection device 220 (e.g., when the statistical detection signal is an F-value signal).

[0039] As indicated by reference numeral 114 in the accompanying drawings, the detection device 220 may (optionally) provide an indication of whether a stable state has been reached. For example, the detection device 220 may provide an indication to the user equipment 230 that the mixing process has reached a stable state. As a specific example, if the detection device 220 has determined that the mixing process has reached a stable state, it may provide an indication to the user equipment 230 associated with monitoring or controlling the mixing process that the mixing process has reached a stable state. As another specific example, if the detection device 220 determines that the mixing process has not yet reached a stable state, it may provide an indication to the user equipment 230 that the mixing process has not yet reached a stable state. In this way, the user of the user equipment 230 can be notified whether the mixing process has reached a stable state (e.g., allowing the user to monitor or adjust the mixing process as needed).

[0040] As another example, the detection device 220 can provide an indication of whether the mixing process has reached a steady state, so that actions can be performed automatically. For example, if the detection device 220 determines that the mixing process is in a steady state, the detection device 220 can provide an indication to one or more other devices, such as devices associated with performing the mixing process (e.g., stopping the mixing process, restarting the mixing process on new raw materials, etc.), or devices associated with performing the next step of the manufacturing process (e.g., starting the next step in the manufacturing process), etc.

[0041] In some implementations, the detection device 220 may provide the user equipment 230 with information about the junction and / or associated with the junction, to enable, for example, visualization of the evolution of the hybrid process via the user equipment 230.

[0042] Figure 1E-Figure 1G This is a graph showing example results associated with the endpoint detection of the hybrid process performed as described above with respect to Example Implementation 100. Figure 1E-Figure 1G This is a graph showing the results when the detection device 220 is configured to perform endpoint detection of the mixing process using the p-value.

[0043] Figure 1E The TSV calculated by detection device 220 throughout the entire iteration of the mixing process is shown. Figure 1E In the example shown, PCA dimensionality reduction is used, and the TSV is calculated based on the product of the standard deviations of the four principal components identified from the spectral data. In this example, the spectral data starting from the pseudo-steady state endpoint (e.g., time step 43) and ending two block sizes later (e.g., at time step 73) is used as the reference block. As the detection device 220 receives spectral data throughout the mixing process, the test block moves. The last test block is shown on the right side of the figure.

[0044] Figure 1F It shows that based on and Figure 1E A graph of the statistical detection signal generated by the associated TSV. Figure 1F In the diagram, the y-axis displays the p-values ​​(with a logarithmic base of 10), and the steady-state threshold is set to -6. In this example, detection device 220 determines at time step 240 that the p-values ​​first satisfy the steady-state threshold, meaning that detection device 220 can determine that the mixing process ends at time step 240 (or shortly thereafter). It is noteworthy that, using the same set of spectral data, conventional techniques would identify the end of the mixing process at time step 42 (e.g., approximately at the end of a pseudo-steady state).

[0045] Figure 1G This is a diagram illustrating the PCA trajectory, which further conveys the ability of the detection device 220 to detect the endpoint of the mixing process. (See diagram.) Figure 1G As shown, corresponding to Figure 1F The data points of the spectral data after the endpoint determined by the detection device 220 (i.e., the spectral data after time step 240) develop into relatively denser clusters, indicating that a stable state of mixing (i.e., mixing uniformity) has been achieved.

[0046] Figure 1H This is a graph illustrating the endpoint detection of the hybrid process performed as described above using the IBSD-based F-value. As mentioned above, a larger F-value (or a smaller p-value) indicates a relatively large difference between the test block and the reference block, which explains... Figure 1F The downward trend of the mixed curve and Figure 1H The upward trend of the mixed curve. Figure 1H In the example shown, the endpoint should be above a predetermined F-value threshold. In this example, the endpoint of the mixing process is determined to be time step 289, which means... Figure 1G The densest clusters of the spectral data shown have reached a stable state. It is noteworthy that, from a practical standpoint, there is no significant difference (e.g., a 3-minute difference) between stopping the mixing process at time step 240 or time step 289 in this example. However, it is important that the detection device 220 does not unreasonably stop the mixing process too early (which would occur with conventional techniques). In some embodiments, as described above, the detection device 220 can be configured to use p-values ​​and F-values ​​associated with determining the endpoint of the mixing process, in which case a more conservative (e.g., later) endpoint identified by the detection device 220 can be used.

[0047] As mentioned above, Figures 1A-1H This is provided as an example. Other examples are also possible, and may differ from those provided. Figures 1A-1H As described.

[0048] Figure 2 This is a diagram of an example environment 200 in which the systems and / or methods described in this paper can be implemented. (See diagram 200 for example environments 200.) Figure 2 As shown, environment 200 may include one or more spectrometers 210-1 to 210-n (n≥1) (collectively referred to herein as spectrometer 210, and individually as spectrometer 210), detection equipment 220, user equipment 230, and network 240. The devices of environment 200 may be interconnected via wired connections, wireless connections, or a combination of wired and wireless connections.

[0049] Spectrometer 210 includes a device capable of performing spectral measurements on a sample (e.g., a sample associated with a manufacturing process). For example, spectrometer 210 may include a benchtop (i.e., non-handheld) spectrometer device that performs spectral analysis (e.g., vibrational spectral analysis, such as near-infrared (NIR) spectral analysis, mid-infrared spectral analysis, Raman spectral analysis, etc.). In some embodiments, spectrometer 210 is capable of providing spectral data acquired by spectrometer 210 for analysis by another device (e.g., detection device 220).

[0050] The detection device 220 includes one or more devices capable of performing one or more operations associated with the detection of the endpoint of the mixing process, as described herein. For example, the detection device 220 may include a server, a server cluster, a computer, a cloud computing device, etc. In some embodiments, the detection device 220 may receive and / or send information to another device in the environment 200, such as spectrometer 210 and / or user equipment 230.

[0051] As described herein, user equipment 230 includes one or more devices capable of receiving, processing, and / or providing information associated with the detection of the endpoint of the hybrid process. For example, user equipment 230 may include communication and computing devices such as desktop computers, mobile phones (e.g., smartphones, cordless phones, etc.), laptop computers, tablet computers, handheld computers, wearable communication devices (e.g., smartwatches, smart glasses, etc.) or similar types of devices.

[0052] Network 240 includes one or more wired and / or wireless networks. For example, network 240 may include cellular networks (e.g., 5G networks, 4G networks, Long Term Evolution (LTE) networks, 3G networks, Code Division Multiple Access (CDMA) networks, etc.), Public Land Mobile Networks (PLMN), Local Area Networks (LAN), Wide Area Networks (WAN), Metropolitan Area Networks (MAN), Telephone Networks (e.g., Public Switched Telephone Network (PSTN)), Private Networks, Self-organizing Networks, Intranets, the Internet, Fiber-based Networks, Cloud Computing Networks, etc., and / or combinations of these or other types of networks.

[0053] Figure 2 The number and arrangement of devices and networks shown are illustrative. In reality, with... Figure 2 Compared to the equipment and / or network shown, there may be additional equipment and / or networks, fewer equipment and / or networks, different equipment and / or networks, or equipment and / or networks with different arrangements. Furthermore, Figure 2 The two or more devices shown can be implemented within a single device, or Figure 2 The single device shown can be implemented as multiple distributed devices. Additionally or alternatively, a collection of devices in environment 200 (e.g., one or more devices) can perform one or more functions described as being performed by another collection of devices in environment 200.

[0054] Figure 3 This is a diagram of example components of device 300, which may correspond to spectrometer 210, detection device 220, and / or user equipment 230. In some embodiments, spectrometer 210, detection device 220, and / or user equipment 230 may include one or more devices 300 and / or one or more components of device 300. Figure 3 As shown, device 300 may include bus 310, processor 320, memory 330, input component 340, output component 350 and communication component 360.

[0055] Bus 310 includes one or more components that enable wired and / or wireless communication between components of device 300. Bus 310 can... Figure 3Two or more components are coupled together, for example via operational coupling, communication coupling, electronic coupling, and / or electrical coupling. Processor 320 includes a central processing unit, graphics processing unit, microprocessor, controller, microcontroller, digital signal processor, field-programmable gate array, application-specific integrated circuit, and / or other types of processing components. Processor 320 is implemented in hardware, firmware, or a combination of hardware and software. In some embodiments, processor 320 includes one or more processors capable of being programmed to perform one or more operations or processes described elsewhere herein.

[0056] Memory 330 includes volatile and / or non-volatile memory. For example, memory 330 may include random access memory (RAM), read-only memory (ROM), hard disk drive, and / or another type of memory (e.g., flash memory, magnetic storage, and / or optical storage). Memory 330 may include internal memory (e.g., RAM, ROM, or hard disk drive) and / or removable memory (e.g., removable via a universal serial bus). Memory 330 may be a non-transitory computer-readable medium. Memory 330 stores information, instructions, and / or software (e.g., one or more software applications) related to the operation of device 300. In some embodiments, memory 330 includes one or more memories coupled to one or more processors (e.g., processor 320), for example, via bus 310.

[0057] Input component 340 enables device 300 to receive input, such as user input and / or sensed input. For example, input component 340 may include a touchscreen, keyboard, keypad, mouse, button, microphone, switch, sensor, GPS sensor, accelerometer, gyroscope, and / or actuator. Output component 350 enables device 300 to provide output, for example, via a display, speaker, and / or light-emitting diode. Communication component 360 enables device 300 to communicate with other devices via wired and / or wireless connections. For example, communication component 360 may include a receiver, transmitter, transceiver, modem, network interface card, and / or antenna.

[0058] Device 300 can perform one or more of the operations or processes described herein. For example, a non-transitory computer-readable medium (e.g., memory 330) can store a set of instructions (e.g., one or more instructions or code) to be executed by processor 320. Processor 320 can execute the set of instructions to perform one or more of the operations or processes described herein. In some embodiments, execution of the set of instructions by one or more processors 320 causes one or more processors 320 and / or device 300 to perform one or more of the operations or processes described herein. In some embodiments, hardwired circuitry is used in place of or in combination with instructions to perform one or more of the operations or processes described herein. Additionally or alternatively, processor 320 may be configured to perform one or more of the operations or processes described herein. Therefore, the embodiments described herein are not limited to any particular combination of hardware circuitry and software.

[0059] Figure 3 The number and arrangement of components shown are provided as an example. Figure 3 Compared to the components shown, device 300 may include additional components, fewer components, different components, or components arranged differently. Additionally or alternatively, a set of components of device 300 (e.g., one or more components) may perform one or more functions described as being performed by another set of components of device 300.

[0060] Figure 4 This is a flowchart of an example process 400 associated with endpoint detection in the mixing process. In some implementations, Figure 4 One or more process blocks are executed by a detection device (e.g., detection device 220). In some implementations, Figure 4 One or more process blocks are performed by another device or group of devices that are separate from or include the detection device, such as a spectrometer (e.g., spectrometer 210) and / or user equipment (e.g., user equipment 230). Additionally or alternatively, Figure 4 One or more process blocks can be executed by one or more components of the device 300, such as processor 320, memory 330, input component 340, output component 350 and / or communication component 360.

[0061] like Figure 4 As shown, process 400 may include receiving spectral data associated with the mixing process (block 410). For example, as described above, the detection device may receive spectral data associated with the mixing process.

[0062] like Figure 4As further shown, process 400 may include identifying a pseudo-steady-state endpoint based on spectral data, the pseudo-steady-state endpoint indicating the end of a pseudo-steady-state associated with the mixing process (box 420). For example, as described above, the detection device may identify a pseudo-steady-state endpoint based on spectral data, the pseudo-steady-state endpoint indicating the end of a pseudo-steady-state associated with the mixing process.

[0063] like Figure 4 As further shown, process 400 may include identifying a reference block and a test block (block 430) from the spectral data based on a pseudo-steady-state endpoint. For example, as described above, the detection device may identify the reference block and the test block from the spectral data based on a pseudo-steady-state endpoint.

[0064] like Figure 4 As further shown, process 400 may include generating a raw detection signal associated with a reference block and a raw detection signal associated with a test block (block 440). For example, as described above, the detection device may generate a raw detection signal associated with a reference block and a raw detection signal associated with a test block.

[0065] like Figure 4 As further shown, process 400 may include generating a statistical detection signal (block 450) based on the original detection signal associated with the reference block and the original detection signal associated with the test block. For example, as described above, the detection device may generate the statistical detection signal based on the original detection signal associated with the reference block and the original detection signal associated with the test block.

[0066] like Figure 4 As further shown, process 400 may include determining whether the mixing process has reached a steady state based on a statistical detection signal (block 460). For example, as described above, the detection device may determine whether the mixing process has reached a steady state based on a statistical detection signal.

[0067] Process 400 may include additional implementations, such as any single implementation or any combination of implementations of one or more other processes described below and / or elsewhere herein.

[0068] In a first embodiment, identifying a pseudo-steady-state endpoint includes performing an F-test using a moving two-block array associated with spectral data to determine a value associated with identifying the pseudo-steady-state endpoint, and to determine that the value satisfies a pseudo-steady-state endpoint threshold.

[0069] In the second embodiment, the F-test is performed alone or in combination with the first embodiment, based on a PCA model generated from the spectra of the moving bi-block intraspectral data.

[0070] In the third embodiment, the F-test is performed, alone or in combination with one or more of the first and second embodiments, based on the entire spectrum from the spectral data within an automatically scaled moving bi-block.

[0071] In the fourth embodiment, the F-test is performed, either alone or in combination with one or more of the first to third embodiments, based on a PCA model generated from historical spectral data.

[0072] In the fifth embodiment, a raw detection signal associated with a reference block is generated, either alone or in combination with one or more of the first to fourth embodiments, based on the TSV calculated based on the reference block.

[0073] In the sixth embodiment, a raw detection signal associated with the test block is generated, either alone or in combination with one or more of the first to fifth embodiments, based on the TSV calculated based on the test block.

[0074] In the seventh embodiment, a raw detection signal associated with a reference block is generated, either alone or in combination with one or more of the first to sixth embodiments, based on the IBSD calculated based on the reference block.

[0075] In the eighth embodiment, a raw detection signal associated with the test block is generated, either alone or in combination with one or more of the first to seventh embodiments, based on the IBSD calculated based on the test block.

[0076] In the ninth embodiment, generating a statistical detection signal, alone or in combination with one or more of the first to eighth embodiments, includes performing an F-test based on a comparison of an original detection signal associated with a reference block and an original detection signal associated with a test block to determine the statistical detection signal.

[0077] In the tenth embodiment, determining whether the mixing process has reached a steady state, alone or in combination with one or more of the first to ninth embodiments, includes determining whether the value of the statistical detection signal satisfies a condition associated with a steady state threshold.

[0078] In the eleventh embodiment, alone or in combination with one or more of the first to tenth embodiments, process 400 includes providing an indication that the mixing process has reached a steady state, based at least in part on the determination that the mixing process has reached a steady state.

[0079] In the twelfth embodiment, alone or in combination with one or more of the first to eleventh embodiments, process 400 includes identifying a junction point based on a pseudo-steady state endpoint, which indicates a transition from an unstable state to a pseudo-steady state.

[0080] although Figure 4 An example block of process 400 is shown, but in some implementations, it is different from... Figure 4 Compared to the blocks shown, process 400 includes additional blocks, fewer blocks, different blocks, or blocks with different arrangements. Alternatively, two or more blocks of process 400 can be executed in parallel.

[0081] The foregoing disclosure provides illustrations and descriptions, but is not intended to be exhaustive or to limit the implementation to the precise forms disclosed. Modifications and variations can be made based on the foregoing disclosure, or modifications and variations can be derived from practical implementation.

[0082] As used herein, the term "component" is intended to be interpreted broadly as hardware, firmware, or a combination of hardware and software. Clearly, the systems and / or methods described herein can be implemented in various forms of hardware, firmware, and / or combinations of hardware and software. The actual dedicated control hardware or software code used to implement these systems and / or methods does not limit these implementations. Therefore, the operation and behavior of the systems and / or methods described herein do not refer to any specific software code, and it should be understood that software and hardware can be used to implement the systems and / or methods based on the description herein.

[0083] As used here, depending on the context, satisfying the threshold can refer to a value that is greater than the threshold, greater than or equal to the threshold, less than the threshold, less than or equal to the threshold, equal to the threshold, not equal to the threshold, etc.

[0084] Even if a particular combination of features is referenced in the claims and / or disclosed in the specification, these combinations are not intended to limit the disclosure of various implementations. In fact, many of these features can be combined in ways not specifically stated in the claims and / or not disclosed in the specification. Although each dependent claim listed below may directly depend on only one claim, the disclosure of various embodiments includes each dependent claim as well as every other claim in the claim set. As used herein, the phrase “at least one” in the list of items refers to any combination of those items, including a single member. For example, “at least one of a, b, or c” is intended to cover any combination of a, b, c, ab, ac, bc, and abc, as well as a plurality of the same items.

[0085] Unless explicitly stated otherwise, the elements, actions, or instructions used herein should not be construed as critical or necessary. Furthermore, as used herein, the article “a” is intended to include one or more items and may be used interchangeably with “one or more.” Furthermore, as used herein, the article “the” is intended to include one or more items associated with the article “the” and may be used interchangeably with “the one or more.” Furthermore, as used herein, the term “set” is intended to include one or more items (e.g., related items, unrelated items, or a combination of related and unrelated items) and may be used interchangeably with “one or more.” When only one item is intended to be used, the phrase “only one” or similar language is used. Furthermore, as used herein, the terms “having,” “possessing,” “having,” etc., are intended to be open-ended terms. Furthermore, the phrase “based on” is intended to mean “at least partially based on,” unless explicitly stated otherwise. Furthermore, as used herein, the term “or” is intended to be included when used in series and may be used interchangeably with “and / or,” unless explicitly stated otherwise (e.g., if used in conjunction with “any” or “only one of them”).

Claims

1. A method for endpoint detection in a mixed process, comprising: The device receives spectral data associated with the mixing process; The device identifies a pseudo-stable state endpoint based on the spectral data. The pseudo-stable state endpoint indicates the end of a pseudo-stable state associated with the mixing process, wherein the pseudo-stable state is a transitional state between an unstable state and a stable state. The device identifies a reference block and a test block from the spectral data based on the pseudo-steady state endpoint, wherein the reference block is a block of the spectral data starting at or near the pseudo-steady state endpoint, and the test block is a block of the spectral data used for comparison with the reference block associated with determining whether the mixing process has reached a steady state; The device generates a raw detection signal associated with the reference block and a raw detection signal associated with the test block; The device generates a statistical detection signal based on the original detection signal associated with the reference block and the original detection signal associated with the test block; as well as The device determines whether the mixing process has reached a stable state based on the statistical detection signal.

2. The method according to claim 1, wherein, Identifying the pseudo-stable state endpoint includes: An F-test is performed using a moving two-block array associated with the spectral data to determine a value associated with identifying the endpoint of the pseudo-steady state; and The value is determined to satisfy the pseudo-steady state endpoint threshold.

3. The method according to claim 2, wherein, The F-test is performed based on a principal component analysis (PCA) model, which is generated based on the spectra from the spectral data within the moving biblock.

4. The method according to claim 2, wherein, The F-test is performed based on automatically scaling the entire spectrum from the spectral data within the moving bi-block.

5. The method according to claim 2, wherein, The F-test is performed based on a principal component analysis (PCA) model, which is generated based on historical spectral data.

6. The method according to claim 1, wherein, The original detection signal associated with the reference block is generated based on the travel space volume (TSV), which is calculated based on the reference block.

7. The method according to claim 1, wherein, The original detection signal associated with the test block is generated based on the travel space volume (TSV), which is calculated based on the test block.

8. The method according to claim 1, wherein, The original detection signal associated with the reference block is generated based on the Block Standard Deviation (IBSD), which is calculated based on the reference block.

9. The method according to claim 1, wherein, The raw detection signal associated with the test block is generated based on the Block Standard Deviation (IBSD), which is calculated based on the test block.

10. The method according to claim 1, wherein, Generating the statistical detection signal includes: An F-test is performed based on a comparison between the original detection signal associated with the reference block and the original detection signal associated with the test block to determine the statistical detection signal.

11. The method according to claim 1, wherein, Determining whether the mixing process has reached a steady state includes: Determine whether the value of the statistical detection signal satisfies the condition associated with the steady-state threshold.

12. The method of claim 1, further comprising providing an indication that the mixing process has reached the steady state, at least in part based on a determination that the mixing process has reached the steady state.

13. The method of claim 1, further comprising identifying junction points based on the pseudo-steady state endpoint, the junction points indicating a transition from an unstable state to the pseudo-steady state.

14. An apparatus for detecting the endpoint of a mixed process, comprising: One or more memory units; and One or more processors coupled to the one or more memories are configured to: Receive spectral data associated with the mixing process; Based on the spectral data, a pseudo-stable state endpoint is identified, which indicates the end of a pseudo-stable state associated with the mixing process, wherein the pseudo-stable state is a transitional state between an unstable state and a stable state. A reference block and a test block are identified from the spectral data based on the pseudo-steady state endpoint, wherein the reference block is a block of the spectral data starting at or near the pseudo-steady state endpoint, and the test block is a block of the spectral data used for comparison with the reference block associated with determining whether the mixing process has reached a steady state. Generate a raw detection signal associated with the reference block and a raw detection signal associated with the test block; A statistical detection signal is generated based on the original detection signal associated with the reference block and the original detection signal associated with the test block; as well as Based on the statistical detection signal, it is determined whether the mixing process has reached a stable state.

15. The device according to claim 14, wherein, In order to identify the pseudo-stable state endpoint, the one or more processors are configured to: An F-test is performed using a moving two-block array associated with the spectral data to determine a value associated with identifying the endpoint of the pseudo-steady state; and The value is determined to satisfy the pseudo-steady state endpoint threshold.

16. The device according to claim 14, wherein, At least one of the original detection signal associated with the reference block or the original detection signal associated with the test block is generated based on travel space volume (TSV) or single block standard deviation (IBSD).

17. The device according to claim 14, wherein, In order to generate the statistical detection signal, the one or more processors are configured to: An F-test is performed based on a comparison between the original detection signal associated with the reference block and the original detection signal associated with the test block to determine the statistical detection signal.

18. The device according to claim 14, wherein, The one or more processors are also configured to identify junctions based on the pseudo-stable state endpoint, which indicates a transition from an unstable state to a pseudo-stable state.

19. The device according to claim 14, wherein, The one or more processors are also configured to provide an indication that the mixing process has reached a stable state, at least in part based on a determination that the mixing process has reached the stable state.

20. A non-transitory computer-readable medium storing an instruction set, the instruction set comprising: One or more instructions, when executed by one or more processors of the device, cause the device to: Receive spectral data associated with the mixing process; Based on the spectral data, a pseudo-stable state endpoint is identified, which indicates the end of a pseudo-stable state associated with the mixing process, wherein the pseudo-stable state is a transitional state between an unstable state and a stable state. Reference blocks and test blocks are identified from the spectral data based on pseudo-steady-state endpoints, wherein the reference block is a block of the spectral data starting at or near the pseudo-steady-state endpoint, and the test block is a block of the spectral data used for comparison with the reference block associated with determining whether the mixing process has reached a steady state. Generate a raw detection signal associated with the reference block and a raw detection signal associated with the test block; A statistical detection signal is generated based on the original detection signal associated with the reference block and the original detection signal associated with the test block; as well as Based on the statistical detection signal, it is determined whether the mixing process has reached a stable state.

Images