A method for establishing a bioaerosol analyzer reliability model

Abstract

The present application relates to the technical field of reliability model, disclose a kind of bioaerosol analyzer reliability model establishing method, for each component unit in bioaerosol analyzer, corresponding model unit is respectively established, and index value assignment is carried out;Obtain the reliability level link and working state of each task mode of bioaerosol analyzer;For each model unit link index assignment is carried out;According to the working state of each component unit in each task mode, for each model unit, working state parameter assignment is carried out, to obtain the reliability parameter of model unit in different task mode;For each model unit, model tree is respectively established;Establish input condition, according to input condition, corresponding reliability model is adaptively established;Wherein, input condition is current task mode and current environmental parameter.The present application can adaptively establish different reliability models according to the task mode change of bioaerosol analyzer.

Description

Technical Field

[0001] This invention relates to the field of reliability modeling technology, and in particular to a method for establishing a reliability model for a bioaerosol analyzer. Background Technology

[0002] A bioaerosol analyzer is an instrument that can monitor the concentration of bioaerosols (i.e., airborne microorganisms, including bacteria, molds, virus clusters, spores, etc.) in the surrounding environment in real time. It can quickly issue an alarm when the concentration of bioaerosols changes abnormally. Its basic working principle is to pass aerosol particles through a high-intensity ultraviolet light beam. Since microbial particles will produce autofluorescence under ultraviolet light, the fluorescence combined with the intensity of scattered light is used to identify whether a single particle belongs to a microorganism. Data statistics and intelligent algorithms are used to determine whether the bioaerosols exceed the normal background range.

[0003] Bioaerosol analyzers are widely used in environmental monitoring, public health and safety (such as pathogen early warning), and biodefense, and their data analysis results are of great significance. The reliability of a bioaerosol analyzer refers to its accuracy, stability, repeatability, and resistance to interference in detecting and identifying airborne biological particles (such as bacteria, viruses, fungi, toxins, etc.). Therefore, the reliability of a bioaerosol analyzer reflects its trouble-free operation and the provision of accurate detection results.

[0004] A reliability model is a mathematical model or framework used to quantify, analyze, or predict the ability of a system, component, or process to perform its intended function under specific conditions and within a specified time. It is widely used in engineering, manufacturing, software development, and risk management to assess and improve system reliability. Establishing a reliability model for a bioaerosol analyzer has significant scientific and engineering value.

[0005] A reliability model consists of two parts: a reliability block diagram and a mathematical formula for calculating the reliability value. Together, they constitute the reliability model of a series system. The reliability block diagram describes the logical reliability relationship between the system and its constituent units; while the calculation formula describes the quantitative reliability relationship between the system and its units.

[0006] There are two main technical shortcomings in the process of establishing the reliability model:

[0007] First, existing reliability model building techniques typically require experienced technicians and experts to participate in the model building process, and the hierarchical structure of the reliability model relies heavily on their involvement. When the operating mode or status of the bioaerosol analyzer changes, it is necessary to organize dedicated personnel to rebuild the model. Therefore, the operation is quite difficult, and it is challenging to make timely evaluations of the operational stability of the bioaerosol analyzer and to provide early warnings and alerts during actual operation.

[0008] Second, each reliability model typically considers a single task mode, building the model based on this mode and setting its parameters. However, bioaerosol analyzers may have multiple different task modes. For example, real-time monitoring mode, sampling-laboratory analysis mode, specific detection mode (such as pathogen-targeted detection), multimodal hybrid mode, or, in actual operation, the bioaerosol analyzer may switch task modes or be adapted to different working environments.

[0009] The operating state of a bioaerosol analyzer differs across different mission phases (for example, different components may exist in both "measurement" and "dormant" states). However, typical reliability models can only assess the reliability of one mission mode. If multiple mission modes need to be evaluated, a separate reliability model must be built for each mode, making the reliability analysis model creation process extremely cumbersome and limiting the applicability of each model. Furthermore, in some cases, a mission may involve different operational phases, each with a different operating mode for the bioaerosol analyzer. Current reliability models are ill-suited for assessing the reliability of multi-phase missions.

[0010] Therefore, the existing methods for establishing reliability models for bioaerosol analyzers have the following drawbacks: each reliability model can only be built manually and is applicable to one task mode. It cannot adaptively establish different reliability models according to changes in the task mode of the bioaerosol analyzer. Summary of the Invention

[0011] The main objective of this invention is to provide a method for establishing a reliability model for a bioaerosol analyzer, which aims to solve the shortcomings of existing technologies where each reliability model establishment method can only be manually built and is applicable to one task mode, and cannot adaptively establish different reliability models according to changes in the task mode of the bioaerosol analyzer.

[0012] To achieve the above objectives, the present invention provides a method for establishing a reliability model for a bioaerosol analyzer, comprising:

[0013] For each component unit in the bioaerosol analyzer, a corresponding model unit is established, and a distinct index value is assigned to each model unit.

[0014] Based on the structured data, unstructured data, and historical sensor time-series data of the bioaerosol analyzer, semantic analysis and knowledge graphs are used to determine the reliable hierarchical links of the bioaerosol analyzer under each task mode, and to obtain the working status of each model unit in the reliable hierarchical links.

[0015] Based on each reliable layer link, a link index value is assigned to each model unit to obtain the link index value of the model unit under different task modes;

[0016] Based on the working state of each component in each task mode, the working state parameters of each model unit are assigned values ​​to obtain the reliability parameters of the model unit in different task modes.

[0017] A model tree is built for each model unit, wherein each model tree stores the link index value and reliability parameters of the model unit under different task modes in a tree structure;

[0018] Establish input conditions, and adaptively build a corresponding reliability model based on the input conditions; wherein the input conditions are the current task mode and the current environment parameters.

[0019] Optionally, the step of determining the reliable hierarchical links of the bioaerosol analyzer for each task mode using semantic analysis and knowledge graphs, based on the structured data, unstructured data, and historical sensor time-series data of the bioaerosol analyzer, includes:

[0020] Acquire structured data, unstructured data, and historical sensor time-series data from the bioaerosol analyzer;

[0021] Extract the constituent units from the structured data to perform semantic analysis on the text information in the unstructured data, and identify the functional dependencies between the components of each constituent unit in the bioaerosol analyzer.

[0022] By performing topological analysis on the unstructured data, the physical connection paths of each component unit in the bioaerosol analyzer are extracted.

[0023] The historical sensor time-series data is analyzed based on the causal discovery algorithm to generate a failure propagation chain;

[0024] Based on the functional dependencies, the physical connection paths, and the failure propagation chains, a knowledge graph is generated with components as nodes and dependencies as edges.

[0025] The physical connection path is mapped to a serial relationship in the reliable hierarchical link, and the redundant path is mapped to a parallel relationship in the reliable hierarchical link.

[0026] Optionally, the step of assigning link index values ​​to each model unit based on each of the reliable layer links to obtain the link index values ​​of the model unit under different task modes includes:

[0027] For each of the model units, establish an index value unit equal to the number of task modes. Each index value unit includes a task mode bit, its own index bit, an upper-level index bit corresponding to the index value of the upper-level index node, and a lower-level index bit corresponding to the index value of the lower-level index node.

[0028] Under different task modes, the first-level index model unit determined from each of the reliable layer links is obtained, and the upper-level index bit of the first-level index model unit is set as the first-level identifier value;

[0029] Based on each reliable hierarchical link, the upper-level index value corresponding to the upper-level index node and the lower-level index value corresponding to the lower-level index node of each component unit are obtained.

[0030] In each index value unit of each model unit, a link index value for each model unit under different task modes is constructed based on the task mode index value, its own index value, the upper layer index value, and the lower layer index value.

[0031] Optionally, the step of assigning working state parameters to each model unit based on the working state of each component unit in each task mode to obtain the reliability parameters of the model unit in different task modes includes:

[0032] The working status of each model unit under different task modes is obtained, as well as several environmental parameter ranges set for each model unit.

[0033] Based on the working status of each model unit in different task modes, reliability values ​​are assigned to each of the environmental parameter ranges set in different task modes.

[0034] Optionally, the step of building a model tree for each model unit includes:

[0035] The first level of the model tree for each model unit is set to the corresponding link index value under different task modes;

[0036] The second level of the model tree for each model unit is set as several environmental parameter ranges corresponding to each link index value;

[0037] The third level of the model tree for each model unit is set as the reliability parameter corresponding to each environmental parameter range for each link index value.

[0038] Optionally, the step of establishing input conditions and adaptively establishing a corresponding reliability model based on the input conditions includes:

[0039] Obtain the current environmental parameter range corresponding to the current environmental parameters monitored by the bioaerosol analyzer;

[0040] Obtain the current task mode index value corresponding to the current task mode of the bioaerosol analyzer;

[0041] The input conditions are established based on the current task mode and the current environment parameter range;

[0042] Based on the input conditions, establish the reliability evaluation hierarchy and reliability parameters corresponding to the current task mode.

[0043] Optionally, the step of extracting the constituent units from the structured data to perform semantic analysis on the text information in the unstructured data and identifying the functional dependencies between the components of each constituent unit in the bioaerosol analyzer includes:

[0044] Extract the set of constituent units using the structured data;

[0045] The set of constituent units is input into the unstructured text and the text is cleaned by using regular expressions to remove semantically irrelevant information and by using syntactic analysis to split complex sentences into simple sentences to obtain the cleaned standard text.

[0046] Retrieve a predefined set of causal verbs:

[0047] Based on the set of constituent units and the set of causal verbs, dependency parsing is performed on the standard text to output each input sentence as a target triple;

[0048] Determine whether the first and third elements of each output target triplet both belong to the elements in the set of constituent units, and whether the second element belongs to the elements in the set of causal verbs. If so, determine the output target triplet as a hierarchical triplet, and determine the functional dependency relationship between the components of each constituent unit in the bioaerosol analyzer based on the hierarchical triplet.

[0049] Optionally, the step of extracting the physical connection paths of each component unit in the bioaerosol analyzer by performing topological analysis on the unstructured data includes:

[0050] The assembly tree is parsed from the assembly design file of the bioaerosol analyzer, and the geometric entities and assembly constraints of all constituent units in the constituent set unit are extracted.

[0051] Collision detection is performed on the bounding boxes of the constituent units to identify directly contacting component pairs;

[0052] Extract connection types from the fit information in the assembly design documents;

[0053] Using the constituent units as nodes and constructing physical connection relationships based on connection types, a hierarchical undirected graph of connections is built for each constituent unit;

[0054] Based on the single connection path in the physical connection relationship, determine the serial connection relationship of component pairs in the hierarchical undirected graph; based on the parallel connection path in the physical connection relationship, determine the parallel connection relationship of component pairs in the hierarchical undirected graph.

[0055] Optionally, the step of analyzing the time-series data in the historical sensor time-series data based on the causal discovery algorithm to generate a failure propagation chain includes:

[0056] Acquire historical time-series data from the bioaerosol analyzer;

[0057] Timestamp alignment of historical time-series data;

[0058] Within a set lag time, causal tests are performed on the fault events based on time-series data. When the correlation between the earlier time-series data and the later fault events reaches a preset correlation threshold, a test conclusion is generated.

[0059] The test results are verified. If the verification is successful, a failure propagation chain is generated based on the time series data and the fault events. The reliability level links are then supplemented based on the failure propagation chain.

[0060] Optionally, the method further includes:

[0061] The environmental parameters and maintenance records of the bioaerosol analyzer under different task modes were obtained;

[0062] From the environmental parameters and the maintenance records, quantify the sampling parameters under different task modes;

[0063] The reliability of each model unit is adjusted based on the sampling frequency and the sample concentration value.

[0064] ;

[0065] Where D represents the degree of vibration damage, i represents the i-th independent vibration frequency band, 1≤i≤M, and M represents the number of vibration frequency bands; Let be the duration of the i-th vibrational frequency band in the vibrational spectrum. The stress of the material in the i-th vibrational frequency band of the vibration spectrum The fatigue life is below;

[0066] when When the damage to the bioaerosol analyzer reaches the set damage value, it indicates that the damage level has reached the set damage value. This triggers an alarm notification.

[0067] when In this case, the reliability of electronic components and mechanical parts are corrected using the following methods:

[0068] ;

[0069] ;

[0070] in, This is a reliability correction value for electronic components. Let e ​​be the initial reliability of the electronic component, and e be the natural constant. The activation energy is given by k, where k is the Boltzmann constant. These are the extreme temperatures in the mission profile. For laboratory temperature testing;

[0071] This is a reliability correction value for mechanical components. For the initial reliability of mechanical components, To design the sampling frequency, This represents the actual sampling frequency in the mission profile. The material wear index, The wear coefficient is the sample concentration.

[0072] In the technical solution of this invention, a corresponding model unit is established for each component unit of the bioaerosol analyzer, and an index value is assigned. Then, according to the reliability hierarchy link of the bioaerosol analyzer under each task mode, a link index value is assigned to each model unit to obtain the link index value of the model unit under different task modes. Thus, after determining the specific task mode, the model units can be automatically connected according to their link index values ​​under the corresponding task mode to build a model block diagram of the reliability model. Simultaneously, based on the specific task mode, the reliability parameters set for the corresponding task mode are obtained, thereby determining the reliability of the bioaerosol analyzer under the corresponding task mode. Therefore, the technical solution of this invention can adaptively build corresponding reliability models according to different single task modes, and can also automatically build reliability models corresponding to multiple task modes. Thus, this invention helps to solve the defect in the prior art where each reliability model establishment method can only be manually built and is applicable to one task mode, and cannot adaptively build different reliability models according to changes in the task mode of the bioaerosol analyzer.

[0073] Furthermore, the present invention can also determine the reliable hierarchical links of the bioaerosol analyzer under each task mode by using semantic analysis and knowledge graphs based on the structured data, unstructured data and historical sensor time-series data of the bioaerosol analyzer. This provides an automatic method for building reliable hierarchical links. Therefore, the reliability model building of the present invention is less dependent on specialized technicians and experts, and improves the intelligence level of building the reliability model. Attached Figure Description

[0074] Figure 1 This is a flowchart of the first embodiment of a method for establishing a reliability model for a bioaerosol analyzer according to the present invention;

[0075] Figure 2 This is a schematic diagram illustrating the establishment of a reliable hierarchical link in this invention;

[0076] Figure 3 This is a schematic diagram illustrating the reliability block diagram established using reliable hierarchical links in this invention.

[0077] Figure 4 This is a schematic diagram of the model tree for each model unit in this invention.

[0078] The objectives, features, and advantages of this invention will be further explained in conjunction with the embodiments and with reference to the accompanying drawings. Detailed Implementation

[0079] It should be understood that the specific embodiments described herein are for illustrative purposes only and are not intended to limit the scope of the invention.

[0080] In the following description, the use of suffixes such as "unit," "component," or "element" to denote elements is solely for the purpose of illustrative purposes and has no specific meaning in itself. Therefore, "unit," "component," or "element" may be used interchangeably.

[0081] Please see Figures 1 to 4 The first embodiment of the present invention provides a method for establishing a reliability model for a bioaerosol analyzer, comprising the following steps:

[0082] Step S10: For each component unit in the bioaerosol analyzer, establish a corresponding model unit and assign distinct index values ​​to each model unit.

[0083] Step S20: Based on the structured data, unstructured data, and historical sensor time-series data of the bioaerosol analyzer, semantic analysis and knowledge graphs are used to determine the reliable hierarchical links of the bioaerosol analyzer under each task mode, and the working status of each model unit in the reliable hierarchical links is obtained.

[0084] Step S30: Assign link index values ​​to each model unit according to each reliable layer link, so as to obtain the link index values ​​of the model unit in different task modes;

[0085] Step S40: Assign working state parameters to each model unit according to the working state of each component unit in each task mode, and obtain the reliability parameters of the model unit in different task modes.

[0086] Step S50: Build a model tree for each model unit, wherein each model tree stores the link index value and reliability parameters of the model unit under different task modes in a tree structure.

[0087] Step S60: Establish input conditions and adaptively establish a corresponding reliability model based on the input conditions; wherein, the input conditions are the current task mode and the current environment parameters.

[0088] In the technical solution of this invention, a corresponding model unit is established for each component unit of the bioaerosol analyzer, and an index value is assigned. Then, according to the reliability hierarchy link of the bioaerosol analyzer under each task mode, a link index value is assigned to each model unit to obtain the link index value of the model unit under different task modes. Thus, after determining the specific task mode, the model units can be automatically connected according to their link index values ​​under the corresponding task mode to build a model block diagram of the reliability model. Simultaneously, based on the specific task mode, the reliability parameters set for the corresponding task mode are obtained, thereby determining the reliability of the bioaerosol analyzer under the corresponding task mode. Therefore, the technical solution of this invention can adaptively build corresponding reliability models according to different single task modes, and can also automatically build reliability models corresponding to multiple task modes. Thus, this invention helps to solve the defect in the prior art where each reliability model establishment method can only be manually built and is applicable to one task mode, and cannot adaptively build different reliability models according to changes in the task mode of the bioaerosol analyzer.

[0089] Furthermore, the present invention can also determine the reliable hierarchical links of the bioaerosol analyzer under each task mode by using semantic analysis and knowledge graphs based on the structured data, unstructured data and historical sensor time-series data of the bioaerosol analyzer. This provides an automatic method for building reliable hierarchical links. Therefore, the reliability model building of the present invention is less dependent on specialized technicians and experts, and improves the intelligence level of building the reliability model.

[0090] Specifically, the reliability model establishment method in this invention is mainly used to establish a reliability model block diagram.

[0091] The reliability hierarchy link includes several evaluation levels connected according to the reliability evaluation level, and each evaluation level includes at least one model unit.

[0092] A bioaerosol analyzer comprises multiple systems, each consisting of multiple components. A typical bioaerosol analyzer includes sampling, pretreatment, detection, and data processing stages, each relying on different components to work together.

[0093] Assigning an index value to each model unit means assigning a unique index value to each model unit. For example, the index value of the sampling pump is 1, and the index value of the filter device is 2.

[0094] The task mode of the bioaerosol analyzer can be set as needed. In this embodiment, the task mode can be: real-time monitoring mode, sampling-laboratory analysis mode, specific detection mode (such as pathogen-targeted detection), multimodal hybrid mode, or one of other working modes. Alternatively, the task mode can be a multi-task mode in stages, which can be real-time monitoring mode, sampling-laboratory analysis mode, specific detection mode (such as pathogen-targeted detection), multimodal hybrid mode, or two or more of other working modes.

[0095] The working order of each component is different in each task mode. For example, the working order in real-time detection mode is: sampling - optical / fluorescence detection - alarm; the working order in sampling-laboratory analysis mode is: sampling - storage - offline analysis; the working order in specific detection mode is: enrichment - target labeling - signal reading - result output; and the working order in multimodal hybrid mode is: parallel real-time + triggered sampling dual-thread asynchronous processing.

[0096] Because the components involved in the work differ in each task mode, the working states of each component are different in each task mode. For example, the working state of each component can be any of the following: standby, low-power operation, and high-power operation. Furthermore, the working order of the components involved in the work will be different in each task mode. For example, in the sampling-lab analysis mode, the sampling component works first, and then the storage component works. Therefore, the order of the reliable hierarchical links formed in each task mode is also different.

[0097] A reliability hierarchy refers to the order of reliability levels determined by the system's functional logic, fault propagation paths, and analysis objectives across different task modes. A reliability hierarchy includes at least one of series or parallel links.

[0098] In each task mode, the reliability of each model unit will vary depending on the corresponding environmental parameters. Therefore, in this invention, a corresponding reliability parameter is set separately for each model unit for each environmental parameter range.

[0099] In a first embodiment of the method for establishing a reliability model for a bioaerosol analyzer based on the present invention, and in a second embodiment of the method for establishing a reliability model for a bioaerosol analyzer based on the present invention, step S20, which involves determining the reliability hierarchy links of the bioaerosol analyzer under each task mode using semantic analysis and knowledge graphs based on the structured data, unstructured data, and historical sensor time-series data of the bioaerosol analyzer, includes:

[0100] Step S21: Obtain the structured data, unstructured data, and historical sensor time-series data of the bioaerosol analyzer;

[0101] Step S22: Extract the constituent units from the structured data to perform semantic analysis on the text information in the unstructured data, and identify the functional dependencies between the components of each constituent unit in the bioaerosol analyzer.

[0102] Step S23: Perform topological analysis on the unstructured data to extract the physical connection paths of each component unit in the bioaerosol analyzer.

[0103] Step S24: Analyze the time series data in the historical sensor time series data based on the causal discovery algorithm to generate a failure propagation chain;

[0104] Step S25: Generate a knowledge graph with components as nodes and dependencies as edges based on the functional dependencies, the physical connection paths, and the failure propagation chains.

[0105] Step S26: Map the physical connection path to a serial relationship in the reliable hierarchical link, and map the redundant path to a parallel relationship in the reliable hierarchical link.

[0106] In a reliability model, the connection logic of each model unit (including series or parallel connection) reflects the physical sequence logic of each component. In the connection logic, the failure of the upstream component will trigger a chain reaction that causes the downstream component to fail. Therefore, determining the connection logic of each model unit is an important part of the reliability model.

[0107] In the prior art, the connection logic of each node in the reliability model is built manually. In this embodiment, a method for automatically building the hierarchical links of the reliability model is provided.

[0108] Specifically, structured data includes at least one of BOM and FMEA tables, while unstructured data includes equipment manuals, fault reports, maintenance logs, and assembly design documents.

[0109] In a first or second embodiment of the method for establishing a reliability model for a bioaerosol analyzer based on the present invention, and in a third embodiment of the method for establishing a reliability model for a bioaerosol analyzer based on the present invention, step S30 includes:

[0110] Step S31: Establish an index value unit for each model unit, with the number of index value units equal to the number of task modes. Each index value unit includes a task mode bit, its own index bit, an upper-level index bit corresponding to the index value of the upper-level index node, and a lower-level index bit corresponding to the index value of the lower-level index node.

[0111] Step S32: Obtain the first-layer index model unit determined from each reliable layer link under different task modes, and set the upper-layer index bit of the first-layer index model unit to the first-layer identifier value;

[0112] Step S33: Based on each reliable hierarchical link, obtain the upper-level index value corresponding to the upper-level index node and the lower-level index value corresponding to the lower-level index node of each component unit.

[0113] Step S34: In each index value unit of each model unit, construct the link index value of each model unit under different task modes based on the task mode index value, its own index value, the upper layer index value, and the lower layer index value.

[0114] Specifically, in each index value unit of each model unit, the task mode index value corresponding to each index value unit is stored in the task mode bit, the self index value corresponding to each model unit is stored in the self index bit, the upper-level index value corresponding to the upper-level index node of each model unit in the corresponding task mode is stored in the upper-level index bit, and the lower-level index value corresponding to the lower-level index node of each model unit in the corresponding task mode is stored in the lower-level index bit, so as to construct the link index value of each model unit in different task modes in each index value unit of each model unit.

[0115] Taking the four task modes (real-time monitoring mode, sampling-laboratory analysis mode, specific detection mode, and multimodal hybrid mode) as an example, each model unit establishes the first index value unit for the real-time monitoring mode, the second index value unit for the sampling-laboratory analysis mode, the third index value unit for the specific detection mode, and the fourth index value unit for the multimodal hybrid mode, so that each model unit has 4 index value units.

[0116] At this point, the model unit corresponds to 4 index value units. Each index value unit includes a link index value with 4 data points. The first bit of the link index value is the task mode bit (used to store the task mode index value corresponding to the task mode), the second bit is the self-index bit (used to store the index value corresponding to the model unit itself, i.e., the self-index value), the third bit is the upper-level index bit (used to store the index value of the upper-level index node corresponding to the current model node in the reliable layer link corresponding to the task mode, i.e., the upper-level index value), and the fourth bit is the lower-level index bit (used to store the index value of the lower-level index node corresponding to the current model node in the reliable layer link corresponding to the task mode, i.e., the lower-level index value).

[0117] Taking the real-time monitoring mode (task mode A) as an example, if the current model unit is a sampling pump (index value 2), its upper-level index node is the model unit corresponding to the particle size selector (index value 5), and its lower-level index node is the model unit corresponding to the flow sensor (index value 8). In this task mode, the link index value corresponding to this index value of the sampling pump model unit is A258. Based on this link index value, the provided link index information is: in the real-time monitoring mode, the upper-level index node of the sampling pump model unit in the reliability evaluation hierarchy is the particle size selector, and the lower-level index node is the flow sensor. Thus, based on the selected task mode, the local reliability evaluation hierarchy is automatically connected, which is easy to understand. Using this method, the reliability evaluation block diagram of the entire task mode can also be connected to establish the framework of the reliability model.

[0118] The first-level identifier can be set as needed; for example, it can be 0.

[0119] In a third embodiment of the method for establishing a reliability model for a bioaerosol analyzer according to the present invention, and in a fourth embodiment of the method for establishing a reliability model for a bioaerosol analyzer according to the present invention, step S40 includes:

[0120] Step S41: Obtain the working status of each model unit under different task modes, and a number of environmental parameter ranges set for each model unit respectively;

[0121] Step S42: Based on the working status of each model unit under different task modes, assign reliability values ​​to each of the environmental parameter ranges set under different task modes.

[0122] Taking the real-time monitoring mode as an example, in the real-time monitoring mode of the bioaerosol analyzer, changes in environmental parameters (such as temperature, humidity, particulate matter concentration, etc.) will significantly affect the reliability of each component.

[0123] Taking the particle size selector in a bioaerosol analyzer as an example, when the particulate matter concentration reaches 500 μg / m³, the risk of blockage at the aerosol inlet and mechanical wear of the particle size selector will accelerate significantly. When the humidity is greater than 80%, water vapor condensation may adhere to particles, changing the particle size cutting efficiency of the particle size selector. Therefore, the reliability parameters of the particle size selector are positively correlated with particle concentration and humidity.

[0124] Taking the sampling pump in a bioaerosol analyzer as an example, the failure rate increases significantly when the temperature is below 0℃ or above 40℃, and its reliability parameters also differ.

[0125] Therefore, in this invention, based on the working state of each model unit under different task modes, reliability values ​​are assigned to each of the environmental parameter ranges set under different task modes.

[0126] In a fourth embodiment of the method for establishing a reliability model for a bioaerosol analyzer according to the present invention, and in a fifth embodiment of the method for establishing a reliability model for a bioaerosol analyzer according to the present invention, step S50 includes:

[0127] Step S51: Set the first level of the model tree of each model unit to the corresponding link index value under different task modes;

[0128] Step S52: Set the second level of the model tree of each model unit to a number of environmental parameter ranges corresponding to each link index value;

[0129] Step S53: Set the third level of the model tree of each model unit to the reliability parameter corresponding to each environmental parameter range of each link index value.

[0130] In this embodiment, a model tree pattern is adopted, in which each model unit stores the link index value corresponding to each task mode and the reliability parameter corresponding to each link index value.

[0131] The advantage of using a model tree approach is that branches can be easily added or deleted. For example, a standard model tree can be pre-stored in a bioaerosol analyzer, and this standard model tree has standard branches. The standard model tree branches take into account the default number of task modes, meeting most reliability assessment requirements.

[0132] When users require more task modes, they can add the necessary branches to the model tree of each model unit to easily import the link index values ​​for the new task modes, as well as the reliability parameters corresponding to the various environmental parameter ranges for each new task mode. Similarly, the link index values ​​and reliability parameters in each model tree branch can be modified or deleted.

[0133] In a fifth embodiment of the method for establishing a reliability model for a bioaerosol analyzer according to the present invention, and in a sixth embodiment of the method for establishing a reliability model for a bioaerosol analyzer according to the present invention, step S60 includes:

[0134] Step S61: Obtain the current environmental parameter range corresponding to the current environmental parameters monitored by the bioaerosol analyzer;

[0135] Step S62: Obtain the current task mode index value corresponding to the current task mode of the bioaerosol analyzer;

[0136] Step S63: Establish the input conditions based on the current task mode and the current environment parameter range;

[0137] Step S64: Based on the input conditions, establish the reliability evaluation hierarchy and reliability parameters corresponding to the current task mode.

[0138] Specifically, the corresponding current task mode index value is determined based on the current task mode of the bioaerosol analyzer;

[0139] Based on the link index value in each of the model units, determine the set of model units corresponding to the current task mode;

[0140] Based on the link index value of each model unit in the model unit set, a reliability evaluation hierarchy architecture corresponding to the current task mode is established.

[0141] Specifically, based on the model units in the model unit set, the model unit whose upper-level index value in the index value unit is the first-level identifier value is determined as the first-level index model unit corresponding to the current task mode; the upper-level index value in the index value unit of each model unit in the model unit set is used to determine the upper-level model unit of each model unit; and the lower-level model unit of each model unit is determined based on the lower-level index value in the index value unit of each model unit in the model unit set, so as to establish a reliability evaluation hierarchy architecture corresponding to the current task mode.

[0142] Furthermore, after obtaining the multiple environmental parameter ranges set in the current task mode, the reliability parameters of the environmental parameter ranges corresponding to the current environmental parameters are established in the model units of the current reliability model.

[0143] Furthermore, the method also includes:

[0144] The current reliability of the bioaerosol analyzer is calculated based on the reliability model corresponding to the current task mode and the reliability parameters of each model unit corresponding to the current environmental parameters.

[0145] According to a second embodiment of the method for establishing a reliability model for a bioaerosol analyzer based on the present invention, and in a seventh embodiment of the method for establishing a reliability model for a bioaerosol analyzer based on the present invention, step S22 includes:

[0146] Step S221: Extract the set of constituent units using the structured data;

[0147] Step S222: Input the set of constituent units into the unstructured text and perform text cleaning, using regular expressions to remove semantically irrelevant information and using syntactic analysis to split complex sentences into simple sentences to obtain the cleaned standard text.

[0148] Step S223: Obtain a preset set of causal verbs:

[0149] Step S224: Based on the set of constituent units and the set of causal verbs, perform dependency parsing on the standard text to output each input sentence as a target triple;

[0150] Step S225: Determine whether the first element and the third element of each output target triplet both belong to the elements in the component unit set, and whether the second element belongs to the elements in the causal verb set. If so, determine the output target triplet as a hierarchical triplet, and determine the functional dependency relationship between the components of each component unit in the bioaerosol analyzer based on the hierarchical triplet.

[0151] The causal verb set includes connective verbs and logical verbs. Connective verbs can directly infer the connection between the first and third elements of the target triple, while logical verbs can determine the pre- and post-causal relationships between the first and second elements. Both connective and causal relationships help determine the hierarchical relationships in a reliable hierarchical link.

[0152] For example, the set of constituent units can be:

[0153] ;

[0154] The set of causal verbs can be:

[0155] ;

[0156] In one specific embodiment, the input sentence in step S224 is "If the fan stops, the temperature of the trigger optical sensor exceeds the limit within 5 minutes", and the output target triplet is <fan, trigger, optical sensor>; in step S225, for example, based on the above target triplet, it can be determined that the fan has a functional dependency on the optical sensor, and thus it can be determined that the fan is a pre-component of the optical sensor in the hierarchical link of the reliability model.

[0157] For example, based on the phrase "the output of the sampling pump is connected to the flow sensor" extracted from the unstructured text, the target triplet can be identified as <sampling pump, connection, flow sensor>, which indicates that the sampling pump is a pre-component of the flow sensor.

[0158] Furthermore, step S22 may also include:

[0159] Step S226: Calculate the relationship weight between upper and lower levels based on the input sentence corresponding to the level triple.

[0160] Specifically, in certain special application scenarios and with specific accuracy requirements, in addition to setting the reliability parameters for each model node, the reliability model can also set relationship weights between nodes at different levels. This relationship weights can be used to determine the probability that the failure of a preceding component will lead to the failure of a subsequent component.

[0161] The formula for calculating relation weights is as follows:

[0162] ;

[0163] ;

[0164] in, For relation weights, It is a pre-defined weight for each causal verb. It is the adverb enhancement coefficient in the input sentence. It is an adverb of intensity; It represents the confidence level of the triple relationship in historical data; n represents the number of times the functional dependency relationship between components of the same triple is accepted in historical data; and N represents the number of times the same triple appears in historical data.

[0165] In the second or seventh embodiment of the method for establishing a reliability model for a bioaerosol analyzer based on the present invention, and in the eighth embodiment of the method for establishing a reliability model for a bioaerosol analyzer based on the present invention, step S23 includes:

[0166] Step S231: Parse the assembly tree from the assembly design file of the bioaerosol analyzer and extract the geometric entities and assembly constraints of all constituent units in the constituent set unit.

[0167] Step S232: Collision detection is performed on the bounding boxes of the constituent units to identify directly contacting component pairs;

[0168] Step S233: Extract the connection type from the fit information in the assembly design document;

[0169] Step S234: Using the constituent units as nodes, construct physical connection relationships based on connection types to build a hierarchical undirected graph of connections for each constituent unit;

[0170] Step S235: Determine the serial connection relationship of component pairs in the hierarchical undirected graph based on the single connection path in the physical connection relationship; determine the parallel connection relationship of component pairs in the hierarchical undirected graph based on the parallel connection path in the physical connection relationship.

[0171] The undirected physical connection skeleton between each component unit can be determined based on the physical connection path in step S23, i.e., the physical connection hierarchy; the directed connection logic between each component unit can be determined based on the functional dependency relationship between components in step S22. Thus, both the hierarchical relationship of the reliable hierarchical link and the hierarchical direction of the reliable hierarchical link can be determined.

[0172] The assembly design file can be a 3D model file (such as an STP file) or a CAD model file.

[0173] Assembly constraints include mating, alignment, and distance constraints.

[0174] Construct a hierarchical undirected graph connecting each component unit. For example, the hierarchical undirected graph could be: sampling port - filter - sampling pump - laser detection cavity - exhaust port.

[0175] In the second, seventh, and eighth embodiments of the method for establishing a reliability model for a bioaerosol analyzer based on the present invention, and in the ninth embodiment of the method for establishing a reliability model for a bioaerosol analyzer based on the present invention, step S24 includes:

[0176] Step S241: Obtain historical time-series data from the bioaerosol analyzer;

[0177] Step S242: Timestamp alignment of historical time series data;

[0178] Step S243: Within a set lag time, perform causal verification on the fault event based on the time series data. When the correlation between the earlier time series data and the later fault event reaches a preset correlation threshold, a verification conclusion is generated.

[0179] Step S244: Verify the test results. If the verification is successful, generate a failure propagation chain based on the time series data and the fault event, and supplement the reliable layer links based on the failure propagation chain.

[0180] In this embodiment, considering functional dependencies and physical connection paths, uncertainties in dynamic behavior, implicit dependencies, and reliability models may be overlooked. Therefore, step S24 is used to generate a failure propagation chain to supplement the missing chain in the reliability hierarchy link.

[0181] For example, in high-temperature environments, the heat dissipation efficiency of the laser (which is functionally dependent on the optical module) decreases, indirectly causing the adjacent, unconnected power supply module to overheat. In this example, the laser is functionally dependent on the optical module; in terms of the physical connection path, the laser is only fixed to the bracket.

[0182] Therefore, based on functional dependencies and physical connection paths, the resulting missing chain is: power failure caused by laser temperature conduction. In this invention, the covariance trend between these two is discovered through time series data in historical sensor time series data.

[0183] For example, in a bioaerosol analyzer, intermittent vibration of the sampling pump (physically connected to the filter) (sensor data) may cause fatigue failure of solder joints on a non-directly connected circuit board. In this example, the identified functional dependency is the relationship between the sampling pump and the filter, and the identified physical connection relationship is the connection between the sampling pump and the filter; the missing chain is: sampling pump vibration causes circuit board failure, resulting in system downtime (time-series correlation needs to be discovered using sensor data).

[0184] Therefore, missing links in the reliable hierarchical link can be supplemented through step S24.

[0185] For example, the process of generating the test conclusion in step S243 can be as follows: test the effect of laser temperature (X) on power supply overheating (Y), with an optimal lag of 2 minutes. If the correlation coefficient reaches 0.8, which is greater than the preset correlation threshold of 0.7, then X and Y are considered to be significantly correlated. The conclusion formed is: laser temperature causes power supply overheating.

[0186] The verification process in step S244 is to eliminate false associations. The verification method is to verify whether the correlation between other prior time-series data and subsequent fault events reaches a preset correlation threshold. If not, the verification is successful, and a failure propagation chain is generated based on the current prior time-series data and fault events. The reliable layer links are supplemented based on the failure propagation chain.

[0187] Specifically, the historical time-series data includes sensor time-series data corresponding to rotational speed, temperature, vibration, current, and air pressure, respectively.

[0188] According to the first to ninth embodiments of the method for establishing a reliability model for a bioaerosol analyzer based on the present invention, in the tenth embodiment of the method for establishing a reliability model for a bioaerosol analyzer based on the present invention, the method further includes:

[0189] Step S70: Obtain the environmental parameters and maintenance records of the bioaerosol analyzer under different task modes;

[0190] Step S80: Quantify the sampling parameters under different task modes from the environmental parameters and the maintenance records;

[0191] Step S90: Correct the reliability of the model units in the reliability model according to the sampling parameters.

[0192] The sampling parameters may include: quantizing the temperature extremes, sampling frequency, vibration spectrum, and sample concentration under different task modes.

[0193] Specifically:

[0194] ;

[0195] Where D represents the degree of vibration damage, i represents the i-th independent vibration frequency band, 1≤i≤M, and M represents the number of vibration frequency bands; Let be the duration of the i-th vibrational frequency band in the vibrational spectrum. The stress of the material in the i-th vibrational frequency band of the vibration spectrum The fatigue life is below;

[0196] when When the damage to the bioaerosol analyzer reaches the set damage value, it indicates that the damage level has reached the set damage value. This triggers an alarm notification.

[0197] when In this case, the reliability of electronic components and mechanical parts are corrected using the following methods:

[0198] ;

[0199] ;

[0200] in, This is a reliability correction value for electronic components. Let e ​​be the initial reliability of the electronic component, and e be the natural constant. The activation energy is given by k, where k is the Boltzmann constant. These are the extreme temperatures in the mission profile. For laboratory temperature testing;

[0201] This is a reliability correction value for mechanical components. For the initial reliability of mechanical components, To design the sampling frequency, This represents the actual sampling frequency in the mission profile. The material wear index, The wear coefficient is the sample concentration.

[0202] The unit of the above temperature is Kelvin, the unit of the above activation energy is eV, and the unit of the above Boltzmann constant is eV / K.

[0203] Through the above description of the embodiments, those skilled in the art can clearly understand that the methods of the above embodiments can be implemented by means of software plus necessary general-purpose hardware platforms. Of course, they can also be implemented by hardware, but in many cases the former is a better implementation method. Based on this understanding, the technical solution of the present invention, or the part that contributes to the prior art, can be embodied in the form of a software product. This computer software product is stored in a computer-readable storage medium (such as ROM / RAM, magnetic disk, optical disk) as described above, and includes several instructions to cause a terminal device to enter the methods described in the various embodiments of the present invention.

[0204] In the description of this specification, references to terms such as "one embodiment," "another embodiment," "other embodiments," or "first embodiment to Xth embodiment," etc., indicate that a specific feature, structure, material, or characteristic described in connection with that embodiment or example is included in at least one embodiment or example of the present invention. In this specification, the illustrative expressions of the above terms do not necessarily refer to the same embodiment or example. Furthermore, the specific features, structures, materials, method steps, or characteristics described may be combined in any suitable manner in one or more embodiments or examples.

[0205] It should be noted that, in this document, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or system that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a process, method, article, or system. Unless otherwise specified, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the process, method, article, or system that includes that element.

[0206] The sequence numbers of the above embodiments of the present invention are for descriptive purposes only and do not represent the superiority or inferiority of the embodiments.

[0207] The above are merely preferred embodiments of the present invention and do not limit the scope of the patent. Any equivalent structural or procedural transformations made based on the description and drawings of the present invention, or direct or indirect applications in other related technical fields, are similarly included within the scope of patent protection of the present invention.

Claims

1. A method for establishing a reliability model for a bioaerosol analyzer, characterized in that, include: For each component unit in the bioaerosol analyzer, a corresponding model unit is established, and a distinct index value is assigned to each model unit. Based on the structured data, unstructured data, and historical sensor time-series data of the bioaerosol analyzer, semantic analysis and knowledge graphs are used to determine the reliable hierarchical links of the bioaerosol analyzer under each task mode, and to obtain the working status of each model unit in the reliable hierarchical links. Based on each reliable layer link, a link index value is assigned to each model unit to obtain the link index value of the model unit under different task modes; Based on the working state of each component in each task mode, the working state parameters of each model unit are assigned values ​​to obtain the reliability parameters of the model unit in different task modes. A model tree is built for each model unit, wherein each model tree stores the link index value and reliability parameters of the model unit under different task modes in a tree structure; Establish input conditions, and adaptively build a corresponding reliability model based on the input conditions; wherein the input conditions are the current task mode and the current environment parameters.

2. The method for establishing a reliability model for a bioaerosol analyzer according to claim 1, characterized in that, The step of determining the reliable hierarchical links of the bioaerosol analyzer for each task mode using semantic analysis and knowledge graphs, based on the structured data, unstructured data, and historical sensor time-series data of the bioaerosol analyzer, includes: Acquire structured data, unstructured data, and historical sensor time-series data from the bioaerosol analyzer; Extract the constituent units from the structured data to perform semantic analysis on the text information in the unstructured data, and identify the functional dependencies between the components of each constituent unit in the bioaerosol analyzer. By performing topological analysis on the unstructured data, the physical connection paths of each component unit in the bioaerosol analyzer are extracted. The historical sensor time-series data is analyzed based on the causal discovery algorithm to generate a failure propagation chain; Based on the functional dependencies, the physical connection paths, and the failure propagation chains, a knowledge graph is generated with components as nodes and dependencies as edges. The physical connection path is mapped to a serial relationship in the reliable hierarchical link, and the redundant path is mapped to a parallel relationship in the reliable hierarchical link.

3. The method for establishing a reliability model for a bioaerosol analyzer according to claim 1, characterized in that, The step of assigning link index values ​​to each model unit based on each reliable layer link to obtain the link index values ​​of the model unit under different task modes includes: For each of the model units, establish an index value unit equal to the number of task modes. Each index value unit includes a task mode bit, its own index bit, an upper-level index bit corresponding to the index value of the upper-level index node, and a lower-level index bit corresponding to the index value of the lower-level index node. Under different task modes, the first-level index model unit determined from each of the reliable layer links is obtained, and the upper-level index bit of the first-level index model unit is set as the first-level identifier value; Based on each reliable hierarchical link, the upper-level index value corresponding to the upper-level index node and the lower-level index value corresponding to the lower-level index node of each component unit are obtained. In each index value unit of each model unit, a link index value for each model unit under different task modes is constructed based on the task mode index value, its own index value, the upper layer index value, and the lower layer index value.

4. The method for establishing a reliability model for a bioaerosol analyzer according to claim 3, characterized in that, The step of assigning working state parameters to each model unit based on the working state of each component unit in each task mode, and obtaining the reliability parameters of the model unit in different task modes, includes: The working status of each model unit under different task modes is obtained, as well as several environmental parameter ranges set for each model unit. Based on the working status of each model unit in different task modes, reliability values ​​are assigned to each of the environmental parameter ranges set in different task modes.

5. The method for establishing a reliability model for a bioaerosol analyzer according to claim 4, characterized in that, The step of building a model tree for each model unit includes: The first level of the model tree for each model unit is set to the corresponding link index value under different task modes; The second level of the model tree for each model unit is set as several environmental parameter ranges corresponding to each link index value; The third level of the model tree for each model unit is set as the reliability parameter corresponding to each environmental parameter range for each link index value.

6. The method for establishing a reliability model for a bioaerosol analyzer according to claim 5, characterized in that, The step of establishing input conditions and adaptively establishing a corresponding reliability model based on the input conditions includes: Obtain the current environmental parameter range corresponding to the current environmental parameters monitored by the bioaerosol analyzer; Obtain the current task mode index value corresponding to the current task mode of the bioaerosol analyzer; The input conditions are established based on the current task mode and the current environment parameter range; Based on the input conditions, establish the reliability evaluation hierarchy and reliability parameters corresponding to the current task mode.

7. The method for establishing a reliability model for a bioaerosol analyzer according to claim 2, characterized in that, The step of extracting the constituent units from the structured data, performing semantic analysis on the text information in the unstructured data, and identifying the functional dependencies between the components of the bioaerosol analyzer includes: Extract the set of constituent units using the structured data; The set of constituent units is input into the unstructured text and the text is cleaned by using regular expressions to remove semantically irrelevant information and by using syntactic analysis to split complex sentences into simple sentences to obtain the cleaned standard text. Retrieve a predefined set of causal verbs: Based on the set of constituent units and the set of causal verbs, dependency parsing is performed on the standard text to output each input sentence as a target triple; Determine whether the first and third elements of each output target triplet both belong to the elements in the set of constituent units, and whether the second element belongs to the elements in the set of causal verbs. If so, determine the output target triplet as a hierarchical triplet, and determine the functional dependency relationship between the components of each constituent unit in the bioaerosol analyzer based on the hierarchical triplet.

8. The method for establishing a reliability model for a bioaerosol analyzer according to claim 2, characterized in that, The step of extracting the physical connection paths of each component unit in the bioaerosol analyzer by performing topological analysis on the unstructured data includes: The assembly tree is parsed from the assembly design file of the bioaerosol analyzer, and the geometric entities and assembly constraints of all constituent units in the assembly unit are extracted. Collision detection is performed on the bounding boxes of the constituent units to identify directly contacting component pairs; Extract connection types from the fit information in the assembly design documents; Using the constituent units as nodes and constructing physical connection relationships based on connection types, a hierarchical undirected graph of connections is built for each constituent unit; Based on the single connection path in the physical connection relationship, determine the serial connection relationship of component pairs in the hierarchical undirected graph; based on the parallel connection path in the physical connection relationship, determine the parallel connection relationship of component pairs in the hierarchical undirected graph.

9. The method for establishing a reliability model for a bioaerosol analyzer according to claim 2, characterized in that, The step of analyzing the time-series data in the historical sensor time-series data based on the causal discovery algorithm to generate a failure propagation chain includes: Acquire historical time-series data from the bioaerosol analyzer; Timestamp alignment of historical time-series data; Within a set lag time, causal tests are performed on the fault events based on time-series data. When the correlation between the earlier time-series data and the later fault events reaches a preset correlation threshold, a test conclusion is generated. The test results are verified. If the verification is successful, a failure propagation chain is generated based on the time series data and the fault events. The reliability level links are then supplemented based on the failure propagation chain.

10. The method for establishing a reliability model for a bioaerosol analyzer according to any one of claims 1 to 9, characterized in that, The method further includes: The environmental parameters and maintenance records of the bioaerosol analyzer under different task modes were obtained; From the environmental parameters and the maintenance records, quantify the sampling parameters under different task modes; The reliability of the model units in the reliability model is corrected based on the sampling parameters. ； Where D represents the degree of vibration damage, i represents the i-th independent vibration frequency band, 1≤i≤M, and M represents the number of vibration frequency bands; Let be the duration of the i-th vibrational frequency band in the vibrational spectrum. The stress of the material in the i-th vibrational frequency band of the vibration spectrum The fatigue life is below; when When the damage to the bioaerosol analyzer reaches the set damage value, it indicates that the damage level has reached the set damage value. This triggers an alarm notification. when In this case, the reliability of electronic components and mechanical parts are corrected using the following methods: ； ； in, This is a reliability correction value for electronic components. Let e ​​be the initial reliability of the electronic component, and e be the natural constant. The activation energy is given by k, where k is the Boltzmann constant. These are the extreme temperatures in the mission profile. For laboratory temperature testing; This is a reliability correction value for mechanical components. For the initial reliability of mechanical components, To design the sampling frequency, This represents the actual sampling frequency in the mission profile. The material wear index, The wear coefficient is the sample concentration.

Images