A method and system for intelligently generating parameters for collecting pollutants in clean environment space

Abstract

The present invention is applicable to the technical field of clean room pollution detection, and in particular, relates to a method and system for intelligently generating parameters for collecting pollutants in a clean environment space. The method comprises generating a test instruction containing a time span, and obtaining an actual pollution parameter matrix at each moment by counting pollution parameters at collection points based on a three-dimensional matrix; selecting different numbers of core points in the actual pollution parameter matrix, and for each number of core points, simulating pollution parameters of other points based on the pollution parameters of the core points to obtain a predicted pollution parameter matrix; comparing the actual pollution parameter matrix with the predicted pollution parameter matrix, and calculating the prediction accuracy of each number; adjusting the collection point activation process in actual application according to the prediction accuracy, and collecting pollution parameters based on the activated collection points; the present invention reduces the number of activations of AMC detectors and reduces the monitoring cost and the possibility of contamination of clean rooms while ensuring data accuracy as much as possible.

Description

Technical Field

[0001] The present invention relates to the technical field of clean room pollution detection, and in particular to a method and system for intelligently generating parameters for collecting pollutants in a clean environment space. Background Art

[0002] Since clean rooms have high requirements for air quality, it is necessary to install a large number of AMC detectors to conduct multi-point AMC concentration detection, thereby achieving accurate monitoring of the clean room environment. In the process of multi-point monitoring, in order to ensure accurate monitoring, each AMC detector is always in working condition to obtain real-time data.

[0003] However, the continuous operation of each AMC detector, on the one hand, causes the consumables in each AMC detector to be continuously consumed, affecting the service life of each AMC detector (the exhaustion of consumables is regarded as the end of the current AMC detector's life, and the consumables need to be replaced after exhaustion), which undoubtedly increases the monitoring cost; on the other hand, the increase in the frequency of replacement of consumables makes the frequency of interaction between the clean room and the outside world increase accordingly, which undoubtedly increases the possibility of the clean room being contaminated by the outside world. Summary of the invention

[0004] The purpose of the present invention is to provide a method and system for intelligently generating parameters for collecting pollutants in clean environment spaces, so as to solve the problem in the above-mentioned background technology that the normalized monitoring of AMC detectors reduces their service life and increases the monitoring cost and pollution risk of clean rooms.

[0005] To achieve the above object, the present invention provides the following technical solutions:

[0006] A method for intelligently generating parameters for collecting pollutants in a clean environment space, the method comprising:

[0007] Acquire an architectural model of the clean room, create collection points based on the architectural model, and synchronously create a three-dimensional matrix according to the positions of the collection points; the elements in the three-dimensional matrix correspond one to one to the collection points;

[0008] Generate test instructions with time span, calculate the pollution parameters at the collection points based on the three-dimensional matrix, and obtain the actual pollution parameter matrix at each moment;

[0009] Different numbers of core points are selected in the actual pollution parameter matrix. For each number of core points, the pollution parameters of other points are simulated based on the pollution parameters of the core points to obtain the predicted pollution parameter matrix.

[0010] Compare the actual pollution parameter matrix with the predicted pollution parameter matrix and calculate the prediction accuracy of each quantity;

[0011] The activation process of the collection points in actual application is adjusted according to the prediction accuracy, and the pollution parameters are collected based on the activated collection points.

[0012] Furthermore, the steps of obtaining the architectural model of the clean room, creating collection points based on the architectural model, and synchronously creating a three-dimensional matrix according to the positions of the collection points include:

[0013] Obtain the BIM model of the clean room and the pollution source distribution information, and fill the pollution sources in the BIM model based on the pollution source distribution information;

[0014] Based on the pollution source, the pollution source space is divided in the BIM model, and the position that does not belong to any pollution source space in the BIM model is used as the background space; the pollution source space is a rectangular space, and the size of the pollution source space corresponds to the pollution source one by one and is a preset value;

[0015] The same number of collection points are evenly set in the pollution source space and the background space;

[0016] Select a corner point in the BIM model, establish a corresponding relationship between the corner point and the origin in the three-dimensional matrix, read other collection points in sequence based on the angle, and establish a corresponding relationship between other collection points and other points in the three-dimensional matrix based on the reading order.

[0017] Furthermore, the step of generating a test instruction with a time span, and obtaining an actual pollution parameter matrix at each moment based on the pollution parameters at the collection points of the three-dimensional matrix statistics includes:

[0018] Receive the test duration input by the administrator as the time span;

[0019] At the preset collection frequency, the pollution parameters are obtained based on the AMC detector installed at the collection point;

[0020] Fill the pollution parameters into the points corresponding to the collection points in the three-dimensional matrix to obtain a three-dimensional matrix containing pollution parameters with time labels, which is called the actual pollution parameter matrix;

[0021] The pollution parameter is an array with at least one element, the element numbers of the array correspond to the pollution types one by one, and the corresponding relationship is preset.

[0022] Furthermore, the step of selecting different numbers of core points in the actual pollution parameter matrix, simulating the pollution parameters of other points based on the pollution parameters of the core points for each number of core points, and obtaining the predicted pollution parameter matrix includes:

[0023] Determine the number of core points and read the actual pollution parameter matrix at each moment in chronological order;

[0024] The number of core points is used as the K value, and the element values ​​in the actual pollution parameter matrix are clustered by K-means to obtain K types of element values ​​and K center points, and the center points are used as core points;

[0025] For each number of core points, the pollution parameters of other points are simulated based on the pollution parameters of the core points to obtain the predicted pollution parameter matrix;

[0026] The simulation process is:

[0027] ; In the formula, For point The predicted value at is a non-core point. The meaning is: connect all core points in pairs, and the resulting line segments are The length of the nearest line segment, is one of the endpoints of the line segment; Indicate point and The projection of the line connecting on the nearest line segment; For point The value at .

[0028] Furthermore, the step of comparing the actual pollution parameter matrix with the predicted pollution parameter matrix and calculating the prediction accuracy of each quantity includes:

[0029] Each core point number is analyzed separately;

[0030] Read the predicted pollution parameter matrix and the actual pollution parameter matrix at all times corresponding to each core point number;

[0031] Compare the predicted pollution parameter matrix and the actual pollution parameter matrix at the same time and calculate the similarity;

[0032] Arrange the similarities in chronological order, calculate the comprehensive similarity, and determine the prediction accuracy;

[0033] Among them, the comprehensive similarity adopts the mean of the similarities, the prediction accuracy is proportional to the comprehensive similarity, and the similarity calculation process is to pair the elements in each row and column position, calculate the difference rate of the two paired elements, and when the difference rate is less than the preset difference rate threshold, mark them as the same elements; calculate the total number of marked identical elements, divide it by the total number of elements in the matrix, and get the similarity.

[0034] Furthermore, the activation process of the collection points in actual application is adjusted according to the prediction accuracy, and the collection of pollution parameters based on the activated collection points includes:

[0035] Query the number of all core points whose prediction accuracy is greater than the preset threshold as the number of core points to be selected;

[0036] Determine the selection probability of each core point according to the inverse ratio of the number of core points to be selected;

[0037] At preset time intervals, the number of core points is selected based on the selection probability, and the activated collection points are updated based on the number of selected core points;

[0038] Among them, the activation process only limits the number of core points, and the specific activation positions are randomly selected.

[0039] The technical solution of the present invention also provides a clean environment space pollutant collection parameter intelligent generation system, the system comprising:

[0040] A collection point creation module is used to obtain the architectural model of the clean room, create collection points based on the architectural model, and synchronously create a three-dimensional matrix according to the positions of the collection points; the elements in the three-dimensional matrix correspond one to one with the collection points;

[0041] The pollution parameter statistics module is used to generate test instructions with time spans, calculate the pollution parameters at the collection points based on the three-dimensional matrix, and obtain the actual pollution parameter matrix at each moment;

[0042] The pollution parameter prediction module is used to select different numbers of core points in the actual pollution parameter matrix, and for each number of core points, simulate the pollution parameters of other points based on the pollution parameters of the core points to obtain a predicted pollution parameter matrix;

[0043] The accuracy calculation module is used to compare the actual pollution parameter matrix with the predicted pollution parameter matrix and calculate the prediction accuracy of each quantity;

[0044] The activation process update module is used to adjust the collection point activation process in actual application according to the prediction accuracy, and collect pollution parameters based on the activated collection points.

[0045] Furthermore, the collection point creation module includes:

[0046] A pollution source filling unit is used to obtain the BIM model of the clean room and the pollution source distribution information, and fill the pollution source in the BIM model based on the pollution source distribution information;

[0047] A space segmentation unit is used to segment a pollution source space in the BIM model based on the pollution source, and use the position in the BIM model that does not belong to any pollution source space as the background space; the pollution source space is a rectangular space, and the size of the pollution source space corresponds to the pollution source one by one and is a preset value;

[0048] A collection point setting unit, used to evenly set the same number of collection points in the pollution source space and the background space;

[0049] The matrix generation unit is used to select a corner point in the BIM model, establish a corresponding relationship between the corner point and the origin in the three-dimensional matrix, read other acquisition points in sequence based on the angle, and establish a corresponding relationship between other acquisition points and other points in the three-dimensional matrix based on the reading order.

[0050] Furthermore, the pollution parameter statistics module includes:

[0051] The duration receiving unit is used to receive the test duration input by the administrator as a time span;

[0052] A pollution parameter acquisition unit, used to acquire pollution parameters based on an AMC detector installed at a collection point at a preset collection frequency;

[0053] A filling execution unit is used to fill the pollution parameters into the points corresponding to the collection points in the three-dimensional matrix, and obtain a three-dimensional matrix containing pollution parameters with time labels, which is called an actual pollution parameter matrix;

[0054] The pollution parameter is an array with at least one element, the element numbers of the array correspond to the pollution types one by one, and the corresponding relationship is preset.

[0055] Furthermore, the pollution parameter prediction module includes:

[0056] The matrix reading unit is used to determine the number of core points and read the actual pollution parameter matrix at each moment in time sequence;

[0057] A clustering unit is used to use the number of core points as K value, perform K-means clustering on the element values ​​in the actual pollution parameter matrix, obtain K class element values ​​and K center points, and use the center points as core points;

[0058] A prediction execution unit is used to simulate the pollution parameters of other points for each number of core points based on the pollution parameters of the core points to obtain a predicted pollution parameter matrix;

[0059] The simulation process is:

[0060] ; In the formula, For point The predicted value at is a non-core point. The meaning is: connect all core points in pairs, and the resulting line segments are The length of the nearest line segment, is one of the endpoints of the line segment; Indicate point and The projection of the line connecting on the nearest line segment; For point The value at .

[0061] Compared with the prior art, the beneficial effects of the present invention are as follows: the present invention first activates all AMC detectors, obtains the collected pollution parameters as standard data, then selects a limited number of data from these standard data as a benchmark, predicts other data, and when the prediction result is accurate enough, records the number as the number of activated AMC detectors in normal use. Therefore, while ensuring the data accuracy as much as possible, the number of activated AMC detectors is reduced, the cost is reduced, the service life is increased in disguise, and the monitoring cost and pollution risk of the clean room are reduced. BRIEF DESCRIPTION OF THE DRAWINGS

[0062] In order to more clearly illustrate the technical solutions in the embodiments of the present invention, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings in the following description are only some embodiments of the present invention.

[0063] Figure 1 A flowchart of a method for intelligently generating clean environment space pollutant collection parameters provided in an embodiment of the present invention.

[0064] Figure 2 A block diagram of the composition of the clean environment space pollutant collection parameter intelligent generation system provided in an embodiment of the present invention. DETAILED DESCRIPTION

[0065] In order to make the purpose, technical solution and advantages of the present invention more clearly understood, the present invention is further described in detail below in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present invention and are not intended to limit the present invention.

[0066] In Example 1, Figure 1 The flowchart of the method for intelligently generating parameters for collecting pollutants in clean environment space provided by the embodiment of the present invention provides a specific implementation process. The method:

[0067] Step S100: Acquire the architectural model of the clean room, create collection points based on the architectural model, and synchronously create a three-dimensional matrix according to the positions of the collection points; the elements in the three-dimensional matrix correspond one to one to the collection points;

[0068] A clean room is a space where chip processing is required. The pollutant concentration in the space is very high. The pollutant is mainly AMC concentration. AMC stands for airborne molecular contaminant (AMC). It is a general term for gaseous pollutants that may have a negative impact on the yield of semiconductor processes, including acids (MA), bases (MB), condensable substances (MC) and dopants (MD). Existing detectors can detect the concentration of AMC.

[0069] Obtain the architectural model of the clean room and create collection points in the architectural model. The collection points are used to install AMC detectors. At the same time, create a matrix based on the positions of the collection points. The row and column positions in the matrix correspond to the AMC detectors one by one and are used to count the data collected by the AMC detectors.

[0070] It should be noted that the operating frequency of all AMC detectors is the same, which means that the AMC detectors acquire data once and perform a matrix filling process once at the same interval.

[0071] In addition, since the distribution of collection points in the clean room is three-dimensional, the corresponding matrix is ​​also three-dimensional. If you want to simplify the analysis process, you can set up planar collection points. For example, only install the AMC detector on the ground. At this time, the three-dimensional matrix will be correspondingly simplified to a two-dimensional matrix.

[0072] Step S200: Generate a test instruction containing a time span, calculate the pollution parameters at the collection points based on the three-dimensional matrix, and obtain the actual pollution parameter matrix at each moment;

[0073] Generate a test instruction containing a time span, which means the test phase starts. The time span indicates the time range in which the data is selected as the test data. Within this time span, the AMC detectors at all collection points need to collect data. The collection process is to collect pollution parameters based on the three-dimensional matrix. At the same time interval, the AMC detector obtains data once and performs a matrix filling process to obtain a matrix containing pollution parameters, which is called the actual pollution parameter matrix.

[0074] Step S300: selecting different numbers of core points in the actual pollution parameter matrix, and for each number of core points, simulating the pollution parameters of other points based on the pollution parameters of the core points to obtain a predicted pollution parameter matrix;

[0075] Some collection points are selected as core points, and then the pollution parameters at the core points are used to simulate the pollution parameters at other locations. This process can be simply described as follows: for an actual pollution parameter matrix, some (number of core points) data are randomly selected from it. With these data, combined with the diffusion simulation algorithm, the data at other points can be inferred. The resulting matrix is ​​called the predicted pollution parameter matrix.

[0076] Among them, the selected points are called core points, and the number of selected points is called the number of core points. The more core points there are, the more accurate the simulation results are. In the extreme case, the total number of elements in the matrix is ​​used as the number of core points. At this time, all data are selected, and the predicted pollution parameter matrix is ​​the same as the actual pollution parameter matrix.

[0077] Furthermore, since an AMC detector is installed at each collection point, the practical significance of the above process is to select several AMC detectors, predict the data in the entire clean room based on the collected data of these AMC detectors, and obtain a predicted pollution parameter matrix. Then, it can be determined whether the results obtained when only these AMC detectors are used are similar to the actual situation (actual pollution parameter matrix).

[0078] Step S400: Compare the actual pollution parameter matrix and the predicted pollution parameter matrix, and calculate the prediction accuracy of each quantity;

[0079] The actual pollution parameter matrix and the predicted pollution parameter matrix are compared. Since the sizes of the two matrices are the same, the comparison process can adopt a conventional similarity calculation scheme. The higher the similarity, the higher the prediction accuracy.

[0080] Step S500: adjusting the collection point activation process in actual application according to the prediction accuracy, and collecting pollution parameters based on the activated collection points;

[0081] Finally, with the prediction accuracy as a reference, the activation process of the collection points is adjusted. The specific adjustment method is: first determine the number of collection points with sufficiently high prediction accuracy. The number of collection points represents the number of activated AMC detectors. The higher the number of collection points, the higher the cost. Correspondingly, the activation time must be shorter. A more common method is to assume that the number of collection points corresponding to the minimum prediction accuracy when the prediction accuracy is high enough is 8, and the total number of collection points is 64. In one hour, the time to activate 8 collection points may be as long as 30 minutes, and the time to activate 64 collection points may be only 2 minutes. Compared with the original solution of activating 64 collection points for one hour, the load becomes very low, which in disguise increases the service life and reduces the cost.

[0082] Regarding step S100, the steps of obtaining the architectural model of the clean room, creating collection points based on the architectural model, and synchronously creating a three-dimensional matrix according to the positions of the collection points include:

[0083] Obtain the BIM model of the clean room and the pollution source distribution information, and fill the pollution sources in the BIM model based on the pollution source distribution information;

[0084] Based on the pollution source, the pollution source space is divided in the BIM model, and the position that does not belong to any pollution source space in the BIM model is used as the background space; the pollution source space is a rectangular space, and the size of the pollution source space corresponds to the pollution source one by one and is a preset value;

[0085] The same number of collection points are evenly set in the pollution source space and the background space;

[0086] Select a corner point in the BIM model, establish a corresponding relationship between the corner point and the origin in the three-dimensional matrix, read other collection points in sequence based on the angle, and establish a corresponding relationship between other collection points and other points in the three-dimensional matrix based on the reading order.

[0087] In an example of the technical solution of the present invention, a BIM model of a clean room and pollution source distribution information are obtained. The BIM model is the basic building structure in the software. According to the distribution position of the actual pollution source, the corresponding components are inserted into the BIM model to obtain a building model containing the pollution source components. Then, based on the BIM model, a rectangular frame is created to intercept the rectangular space, which is called the pollution source space. The pollution source space is eliminated in the building model containing the pollution source components, and the remaining space is called the background space. The same number of collection points are evenly set in the pollution source space and the background space. Then, based on the positional relationship between the collection points, a three-dimensional matrix is ​​created, and the one-to-one correspondence between the collection points and the elements in the matrix is ​​determined.

[0088] It should be noted that the meaning of uniform setting is that the distances between each collection point are the same, which can be achieved with the help of a grid. Since the pollution source space is smaller than the background space, when the number of collection points in the pollution source space is the same as the number of collection points in the background space, the collection density in the pollution source space will be higher, which is consistent with the actual situation.

[0089] Regarding step S200, the step of generating a test instruction containing a time span, and obtaining an actual pollution parameter matrix at each moment based on the pollution parameters at the collection points of the three-dimensional matrix statistics includes:

[0090] Receive the test duration input by the administrator as the time span;

[0091] At the preset collection frequency, the pollution parameters are obtained based on the AMC detector installed at the collection point;

[0092] Fill the pollution parameters into the points corresponding to the collection points in the three-dimensional matrix to obtain a three-dimensional matrix containing pollution parameters with time labels, which is called the actual pollution parameter matrix;

[0093] The pollution parameter is an array with at least one element, the element numbers of the array correspond to the pollution types one by one, and the corresponding relationship is preset.

[0094] An AMC detector is installed at the collection point. At the preset collection frequency, the pollution parameters are obtained and inserted into the three-dimensional matrix to obtain the actual pollution parameter matrix. The administrator enters a duration as the time span, and all actual pollution parameter matrices within the time span are used as data to be analyzed.

[0095] It is worth mentioning that the pollution parameter may include data of multiple pollution types. Therefore, the pollution parameter may be an array, and each element of the array corresponds to a pollution type.

[0096] Regarding step S300, the step of selecting different numbers of core points in the actual pollution parameter matrix, simulating the pollution parameters of other points based on the pollution parameters of the core points for each number of core points, and obtaining the predicted pollution parameter matrix includes:

[0097] Determine the number of core points and read the actual pollution parameter matrix at each moment in chronological order;

[0098] The number of core points is used as the K value, and the element values ​​in the actual pollution parameter matrix are clustered by K-means to obtain K types of element values ​​and K center points, and the center points are used as core points;

[0099] For each number of core points, the pollution parameters of other points are simulated based on the pollution parameters of the core points to obtain the predicted pollution parameter matrix.

[0100] For the actual pollution parameter matrix, the clustering process is not difficult. First, determine the number of categories, that is, the number of core points, and use it as the K value to perform K-means clustering on each element in the actual pollution parameter matrix. You can get K center points, which are the core points; different numbers of categories correspond to different K center points.

[0101] Then, for each number of core points, the pollution parameters of other points are simulated based on the pollution parameters of the core points. The resulting matrix is ​​called the predicted pollution parameter matrix. One of the simulation processes is described as follows:

[0102] ; In the formula, For point The predicted value at is a non-core point. The meaning is: connect all core points in pairs, and the resulting line segments are The length of the nearest line segment, is one of the endpoints of the line segment; Indicate point and The projection of the line connecting on the nearest line segment; For point The value at .

[0103] The above simulation process is relatively simple. The core points are connected in pairs to obtain multiple line segments. For any point, the nearest line segment is queried (using the distance from the point to the line segment), and the point is simulated by this line segment. The specific simulation principle is to select a core point in the line segment, connect the point and the core point, calculate the projection, calculate the ratio of the projection to the line segment, and then multiply it by the value at the core point to get the predicted value; among them, the value at the core point is a known value.

[0104] Regarding step S400, the step of comparing the actual pollution parameter matrix with the predicted pollution parameter matrix and calculating the prediction accuracy of each quantity includes:

[0105] Each core point number is analyzed separately;

[0106] Read the predicted pollution parameter matrix and the actual pollution parameter matrix at all times corresponding to each core point number;

[0107] Compare the predicted pollution parameter matrix and the actual pollution parameter matrix at the same time and calculate the similarity;

[0108] Arrange the similarities in chronological order, calculate the comprehensive similarity, and determine the prediction accuracy;

[0109] Among them, the comprehensive similarity adopts the mean of the similarities, the prediction accuracy is proportional to the comprehensive similarity, and the similarity calculation process is to pair the elements in each row and column position, calculate the difference rate of the two paired elements, and when the difference rate is less than the preset difference rate threshold, mark them as the same elements; calculate the total number of marked identical elements, divide it by the total number of elements in the matrix, and get the similarity.

[0110] For each number of core points, each collection moment in the test period corresponds to a predicted pollution parameter matrix and an actual pollution parameter matrix. The predicted pollution parameter matrix and the actual pollution parameter matrix at the same collection moment are compared, the similarity is calculated, and then the similarity is arranged in chronological order, and then the comprehensive similarity is calculated to obtain the prediction accuracy rate.

[0111] Regarding step S500, the activation process of the collection points in actual application is adjusted according to the prediction accuracy, and the collection of pollution parameters based on the activated collection points includes:

[0112] Query the number of all core points whose prediction accuracy is greater than the preset threshold as the number of core points to be selected;

[0113] Determine the selection probability of each core point according to the inverse ratio of the number of core points to be selected;

[0114] At preset time intervals, the number of core points is selected based on the selection probability, and the activated collection points are updated based on the number of selected core points;

[0115] Among them, the activation process only limits the number of core points, and the specific activation positions are randomly selected.

[0116] The prediction accuracy is compared with the preset threshold. When the prediction accuracy is greater than the preset threshold, the corresponding number of core points is used as the number of core points to be selected. The selection probability of each core point is determined according to the inverse proportion of the number of core points to be selected. At this time, a duration is no longer allocated to each number of core points, but a random application scheme based on the selection probability is used. At regular intervals, a collection point update behavior is triggered, and the number of activated collection points is determined based on the selection probability, and then the number of collection points is randomly selected from all collection points for activation.

[0117] As a preferred embodiment of the technical solution of the present invention, Figure 2 This is a block diagram of the composition of the clean environment space pollutant collection parameter intelligent generation system provided by an embodiment of the present invention. The clean environment space pollutant collection parameter intelligent generation system 10 includes:

[0118] The collection point creation module 11 is used to obtain the architectural model of the clean room, create collection points based on the architectural model, and synchronously create a three-dimensional matrix according to the positions of the collection points; the elements in the three-dimensional matrix correspond to the collection points one by one;

[0119] The pollution parameter statistics module 12 is used to generate a test instruction containing a time span, and to obtain the actual pollution parameter matrix at each moment based on the pollution parameters at the collection points of the three-dimensional matrix statistics;

[0120] The pollution parameter prediction module 13 is used to select different numbers of core points in the actual pollution parameter matrix, and for each number of core points, simulate the pollution parameters of other points based on the pollution parameters of the core points to obtain a predicted pollution parameter matrix;

[0121] The accuracy calculation module 14 is used to compare the actual pollution parameter matrix with the predicted pollution parameter matrix and calculate the prediction accuracy of each quantity;

[0122] The activation process updating module 15 is used to adjust the collection point activation process in actual application according to the prediction accuracy, and collect pollution parameters based on the activated collection points.

[0123] Furthermore, the collection point creation module 11 includes:

[0124] A pollution source filling unit is used to obtain the BIM model of the clean room and the pollution source distribution information, and fill the pollution source in the BIM model based on the pollution source distribution information;

[0125] A space segmentation unit is used to segment a pollution source space in the BIM model based on the pollution source, and use the position in the BIM model that does not belong to any pollution source space as the background space; the pollution source space is a rectangular space, and the size of the pollution source space corresponds to the pollution source one by one and is a preset value;

[0126] A collection point setting unit, used to evenly set the same number of collection points in the pollution source space and the background space;

[0127] The matrix generation unit is used to select a corner point in the BIM model, establish a corresponding relationship between the corner point and the origin in the three-dimensional matrix, read other acquisition points in sequence based on the angle, and establish a corresponding relationship between other acquisition points and other points in the three-dimensional matrix based on the reading order.

[0128] Specifically, the pollution parameter statistics module 12 includes:

[0129] A duration receiving unit, used for receiving the test duration input by the administrator as a time span;

[0130] A pollution parameter acquisition unit, used to acquire pollution parameters based on an AMC detector installed at a collection point at a preset collection frequency;

[0131] A filling execution unit is used to fill the pollution parameters into the points corresponding to the collection points in the three-dimensional matrix, and obtain a three-dimensional matrix containing pollution parameters with time labels, which is called an actual pollution parameter matrix;

[0132] The pollution parameter is an array with at least one element, the element numbers of the array correspond to the pollution types one by one, and the corresponding relationship is preset.

[0133] In addition, the pollution parameter prediction module 13 includes:

[0134] The matrix reading unit is used to determine the number of core points and read the actual pollution parameter matrix at each moment in time sequence;

[0135] A clustering unit is used to use the number of core points as K value, perform K-means clustering on the element values ​​in the actual pollution parameter matrix, obtain K class element values ​​and K center points, and use the center points as core points;

[0136] A prediction execution unit is used to simulate the pollution parameters of other points for each number of core points based on the pollution parameters of the core points to obtain a predicted pollution parameter matrix;

[0137] The simulation process is:

[0138] ; In the formula, For point The predicted value at is a non-core point. The meaning is: connect all core points in pairs, and the resulting line segments are The length of the nearest line segment, is one of the endpoints of the line segment; Indicate point and The projection of the line connecting on the nearest line segment; For point The value at .

[0139] The above are only preferred embodiments of the present invention, and are not intended to limit the patent scope of the present invention. Any equivalent structure or equivalent process transformation made using the contents of the present invention specification and drawings, or directly or indirectly applied in other related technical fields, are also included in the patent protection scope of the present invention.

Claims

1. A method for intelligently generating parameters for collecting pollutants in clean environment space, characterized in that: The method comprises: Acquire an architectural model of the clean room, create collection points based on the architectural model, and synchronously create a three-dimensional matrix according to the positions of the collection points; the elements in the three-dimensional matrix correspond one to one to the collection points; Generate test instructions with time span, calculate the pollution parameters at the collection points based on the three-dimensional matrix, and obtain the actual pollution parameter matrix at each moment; Different numbers of core points are selected in the actual pollution parameter matrix. For each number of core points, the pollution parameters of other points are simulated based on the pollution parameters of the core points to obtain the predicted pollution parameter matrix. Compare the actual pollution parameter matrix with the predicted pollution parameter matrix and calculate the prediction accuracy of each quantity; adjusting the collection point activation process in actual application according to the prediction accuracy, and collecting pollution parameters based on the activated collection points; The step of generating a test instruction with a time span, and obtaining an actual pollution parameter matrix at each moment based on the pollution parameters at the collection points of the three-dimensional matrix statistics comprises: Receive the test duration input by the administrator as the time span; At the preset collection frequency, the pollution parameters are obtained based on the AMC detector installed at the collection point; Fill the pollution parameters into the points corresponding to the collection points in the three-dimensional matrix to obtain a three-dimensional matrix containing pollution parameters with time labels, which is called the actual pollution parameter matrix; The pollution parameter is an array with at least one element, the element number of the array corresponds to the pollution type one by one and the corresponding relationship is preset; The step of selecting different numbers of core points in the actual pollution parameter matrix, simulating the pollution parameters of other points based on the pollution parameters of the core points for each number of core points, and obtaining the predicted pollution parameter matrix includes: Determine the number of core points and read the actual pollution parameter matrix at each moment in chronological order; The number of core points is used as the K value, and the element values ​​in the actual pollution parameter matrix are clustered by K-means to obtain K types of element values ​​and K center points, and the center points are used as core points; For each number of core points, the pollution parameters of other points are simulated based on the pollution parameters of the core points to obtain the predicted pollution parameter matrix; The simulation process is: ; In the formula, For point The predicted value at is a non-core point. The meaning is: connect all core points in pairs, and the resulting line segments are The length of the nearest line segment, is one of the endpoints of the line segment; Indicate point and The projection of the line connecting on the nearest line segment; For point The value at The step of comparing the actual pollution parameter matrix with the predicted pollution parameter matrix and calculating the prediction accuracy of each quantity includes: Each core point number is analyzed separately; Read the predicted pollution parameter matrix and the actual pollution parameter matrix at all times corresponding to each core point number; Compare the predicted pollution parameter matrix and the actual pollution parameter matrix at the same time and calculate the similarity; Arrange the similarities in chronological order, calculate the comprehensive similarity, and determine the prediction accuracy; Among them, the comprehensive similarity adopts the mean of the similarities, the prediction accuracy is proportional to the comprehensive similarity, and the similarity calculation process is: pair the elements in each row and column position, calculate the difference rate of the two paired elements, and when the difference rate is less than the preset difference rate threshold, mark them as the same elements; calculate the total number of marked identical elements, divide it by the total number of elements in the matrix, and get the similarity.

2. The method for intelligently generating parameters for collecting pollutants in clean environment space according to claim 1 is characterized in that: The steps of obtaining the architectural model of the clean room, creating collection points based on the architectural model, and synchronously creating a three-dimensional matrix according to the positions of the collection points include: Obtain the BIM model of the clean room and the pollution source distribution information, and fill the pollution sources in the BIM model based on the pollution source distribution information; Based on the pollution source, the pollution source space is divided in the BIM model, and the position that does not belong to any pollution source space in the BIM model is used as the background space; the pollution source space is a rectangular space, and the size of the pollution source space corresponds to the pollution source one by one and is a preset value; The same number of collection points are evenly set in the pollution source space and the background space; Select a corner point in the BIM model, establish a corresponding relationship between the corner point and the origin in the three-dimensional matrix, read other collection points in sequence based on the angle, and establish a corresponding relationship between other collection points and other points in the three-dimensional matrix based on the reading order.

3. The method for intelligently generating parameters for collecting pollutants in clean environment space according to claim 1 is characterized in that: The process of adjusting the collection point activation in actual application according to the prediction accuracy, collecting pollution parameters based on the activated collection points includes: Query the number of all core points whose prediction accuracy is greater than the preset threshold as the number of core points to be selected; Determine the selection probability of each core point according to the inverse ratio of the number of core points to be selected; At preset time intervals, the number of core points is selected based on the selection probability, and the activated collection points are updated based on the number of selected core points; Among them, the activation process only limits the number of core points, and the specific activation positions are randomly selected.

4. A clean environment space pollutant collection parameter intelligent generation system, characterized in that: The system comprises: A collection point creation module is used to obtain the architectural model of the clean room, create collection points based on the architectural model, and synchronously create a three-dimensional matrix according to the positions of the collection points; the elements in the three-dimensional matrix correspond one to one with the collection points; The pollution parameter statistics module is used to generate test instructions with time spans, calculate the pollution parameters at the collection points based on the three-dimensional matrix, and obtain the actual pollution parameter matrix at each moment; The pollution parameter prediction module is used to select different numbers of core points in the actual pollution parameter matrix, and for each number of core points, simulate the pollution parameters of other points based on the pollution parameters of the core points to obtain a predicted pollution parameter matrix; The accuracy calculation module is used to compare the actual pollution parameter matrix with the predicted pollution parameter matrix and calculate the prediction accuracy of each quantity; An activation process update module, used for adjusting the collection point activation process in actual application according to the prediction accuracy, and collecting pollution parameters based on the activated collection points; The pollution parameter statistics module includes: The duration receiving unit is used to receive the test duration input by the administrator as a time span; A pollution parameter acquisition unit, used to acquire pollution parameters based on an AMC detector installed at a collection point at a preset collection frequency; A filling execution unit is used to fill the pollution parameters into the points corresponding to the collection points in the three-dimensional matrix, and obtain a three-dimensional matrix containing pollution parameters with time labels, which is called an actual pollution parameter matrix; The pollution parameter is an array with at least one element, the element number of the array corresponds to the pollution type one by one and the corresponding relationship is preset; The pollution parameter prediction module includes: The matrix reading unit is used to determine the number of core points and read the actual pollution parameter matrix at each moment in time sequence; A clustering unit is used to use the number of core points as K value, perform K-means clustering on the element values ​​in the actual pollution parameter matrix, obtain K class element values ​​and K center points, and use the center points as core points; A prediction execution unit is used to simulate the pollution parameters of other points for each number of core points based on the pollution parameters of the core points to obtain a predicted pollution parameter matrix; The simulation process is: ; In the formula, For point The predicted value at is a non-core point. The meaning is: connect all core points in pairs, and the resulting line segments are The length of the nearest line segment, is one of the endpoints of the line segment; Indicate point and The projection of the line connecting on the nearest line segment; For point The value at The content of comparing the actual pollution parameter matrix with the predicted pollution parameter matrix and calculating the prediction accuracy of each quantity includes: Each core point number is analyzed separately; Read the predicted pollution parameter matrix and the actual pollution parameter matrix at all times corresponding to each core point number; Compare the predicted pollution parameter matrix and the actual pollution parameter matrix at the same time and calculate the similarity; Arrange the similarities in chronological order, calculate the comprehensive similarity, and determine the prediction accuracy; Among them, the comprehensive similarity adopts the mean of the similarities, the prediction accuracy is proportional to the comprehensive similarity, and the similarity calculation process is: pair the elements in each row and column position, calculate the difference rate of the two paired elements, and when the difference rate is less than the preset difference rate threshold, mark them as the same elements; calculate the total number of marked identical elements, divide it by the total number of elements in the matrix, and get the similarity.

5. The clean environment space pollutant collection parameter intelligent generation system according to claim 4 is characterized in that: The collection point creation module includes: A pollution source filling unit is used to obtain the BIM model of the clean room and the pollution source distribution information, and fill the pollution source in the BIM model based on the pollution source distribution information; A space segmentation unit is used to segment a pollution source space in the BIM model based on the pollution source, and use the position in the BIM model that does not belong to any pollution source space as the background space; the pollution source space is a rectangular space, and the size of the pollution source space corresponds to the pollution source one by one and is a preset value; A collection point setting unit, used to evenly set the same number of collection points in the pollution source space and the background space; The matrix generation unit is used to select a corner point in the BIM model, establish a corresponding relationship between the corner point and the origin in the three-dimensional matrix, read other acquisition points in sequence based on the angle, and establish a corresponding relationship between other acquisition points and other points in the three-dimensional matrix based on the reading order.

Images