Redundancy configuration method and device of reaction kettle array and computer equipment

Abstract

The application discloses a redundancy configuration method and device of a reaction kettle array and computer equipment, and relates to the field of chemical equipment automation control. The redundancy configuration method of the reaction kettle array comprises determining the health degree score of each reaction kettle before the reaction kettle array produces a target sample; determining the first failure probability of each reaction kettle according to the difference in the health degree score of each reaction kettle through a distribution function; determining the second failure probability of the simultaneous failure of multiple reaction kettles in the reaction kettle array according to the first failure probability of each reaction kettle; substituting the second failure probability into a total cost function constructed in advance to obtain a target value when the function value of the total cost function is the minimum, and the target value is used to indicate the number of failures of the reaction kettles; and awakening a corresponding number of standby kettles according to the target value to replace the failed reaction kettles. Through the above method, the property loss caused by the failure of the reaction kettles is reduced.

Description

Technical Field

[0001] This application relates to the field of automated control of chemical equipment, and in particular to a method, apparatus and computer equipment for redundant configuration of a reactor array. Background Technology

[0002] A reactor array is a system composed of several reactors used in the flexible manufacturing of high-end chemicals. During the use of a reactor array, if a reactor malfunctions, a backup reactor must be activated promptly to replace the faulty one.

[0003] If the configuration of the spare reactors is completed during the factory design phase, on the one hand, the number of spare reactors cannot be updated in real time according to the usage after the reactor array is put into production; on the other hand, the usage of each reactor after it is put into production will vary, resulting in different failure rates for each reactor. Therefore, the number of spare reactors cannot be reasonably configured according to the failure rate of each reactor, which leads to increased production costs and high property losses. Summary of the Invention

[0004] To address the shortcomings of existing technologies, the purpose of this application is to provide a method, apparatus, and computer equipment for redundant configuration of reactor arrays, which can reduce property losses when reactors malfunction.

[0005] To achieve the above objectives, this application adopts the following technical solution:

[0006] In a first aspect, this application provides a method for redundant configuration of a reactor array, comprising: determining a health score for each reactor in the reactor array before producing a target sample; determining a first failure probability for each reactor based on the difference in health scores using a distribution function; wherein the first failure probability of each reactor varies due to state differences, including at least one of differences in the number of times the reactor is used, differences in the lifespan of components, and differences in the replacement time of components; determining a second failure probability of simultaneous failure of multiple reactors in the reactor array based on the first failure probability of each reactor; substituting the second failure probability into a pre-constructed total cost function to obtain a target value that minimizes the function value of the total cost function, the target value indicating the number of reactor failures; wherein the total cost function includes a risk cost function indicating the maximum expected loss due to the scrapping of the target sample caused by the number of failures exceeding the total number of spare reactors in the reactor array; and activating a corresponding number of spare reactors according to the target value to replace the failed reactors.

[0007] In some embodiments, the total cost function further includes a holding cost function and a switching loss function; the holding cost function is used to indicate the power-on preheating cost and depreciation cost of waking up the standby vessel; the switching loss function is used to indicate the material cleaning loss of the malfunctioning reactor.

[0008] In some embodiments, the total cost function The following relationship must be satisfied:

[0009] ;

[0010] In the formula, This represents the power-on preheating cost and depreciation cost of a spare reactor. E represents the material washing loss of a reactor; E represents the expected cost. Indicates the expected number of switching operations.

[0011] In some embodiments, the maximum expected loss is positively correlated with the material value of the target sample produced.

[0012] In some embodiments, the risk cost function The following relationship must be satisfied:

[0013] ;

[0014] In the formula, α Indicates the confidence interval. S Indicates the total number of spare reactors. C batch Let P represent the value of the material and P represent the second failure probability. K=k This indicates the number of reactor failures, where M represents the total number of reactors currently in operation. This represents the risk aversion coefficient.

[0015] In some embodiments, a second failure probability of simultaneous failure of multiple reactors in the reactor array is determined based on the Poisson binomial distribution of the first failure probability of each reactor. This includes calculating the distribution of the first failure probability using a Poisson binomial distribution model and solving the problem using a numerical algorithm based on recursion or discrete Fourier transform to obtain the second failure probability of simultaneous failure of multiple reactors in the reactor array.

[0016] In some embodiments, the health score of each reactor in the reactor array before producing the target sample is determined, including sampling process data of the reactor array during the production of the target sample; extracting a first time-series feature dataset from the process data and inputting the first time-series feature dataset into a preset time-series neural network model to obtain a second time-series feature dataset after dimensionality reduction; determining the target feature vector dataset at the current moment based on the historical production dataset of the target sample and the second time-series feature dataset, wherein the target feature vector dataset is used to indicate the residual between the historical production dataset and the process data at the current moment; and determining the health score of each reactor in the reactor array based on the weighted Mahalanobis distance of the target feature vector dataset.

[0017] Secondly, this application provides a redundant configuration device for a reactor array, comprising a data acquisition module, a data processing module, and an execution module. The data acquisition module is used to determine the health score of each reactor in the reactor array before producing the target sample. The data processing module is used to determine the first failure probability of each reactor based on the difference in the health scores of each reactor using a distribution function. The first failure probability of each reactor varies due to state differences, which include at least one of the following: differences in the number of times the reactor is used, differences in the lifespan of parts, and differences in the replacement time of parts. Based on the Poisson binomial distribution of the first failure probability of each reactor, a second failure probability of multiple reactors in the reactor array failing simultaneously is determined. The second failure probability is substituted into a pre-constructed total cost function to obtain a target value that minimizes the function value of the total cost function. The target value is used to indicate the number of reactor failures. The total cost function includes a risk cost function, which indicates the maximum expected loss caused by the scrapping of the target sample when the number of failures exceeds the total number of spare reactors in the reactor array. The execution module is used to wake up a corresponding number of spare reactors according to the target value to replace the failed reactors.

[0018] Thirdly, this application provides a computer device including a memory and a processor, wherein the memory stores a computer program, and when the computer program is executed by the processor, it performs any of the above-described redundancy configuration methods.

[0019] Fourthly, this application provides a computer-readable storage medium storing program instructions that, when executed on one or more processors, implement any of the aforementioned redundancy configuration methods.

[0020] The redundant configuration method for reactor arrays provided in this application determines the health score of each reactor in the reactor array before producing the target sample. Based on the health score, it determines the first failure probability of each reactor and the second failure probability of multiple reactors failing simultaneously. The second failure probability is then incorporated into a pre-constructed total cost function. Based on various costs when switching to a backup reactor, a target value with the lowest total cost is obtained. The corresponding number of backup reactors are then activated and preheated according to the target value, allowing the backup reactors to replace the failed reactors. After the reactor array is put into production, the number of backup reactors activated is dynamically adjusted based on the individual differences of each reactor, thereby reducing property losses when reactors fail. Attached Figure Description

[0021] Figure 1 This is a schematic diagram of the redundant configuration device of the reactor array in the embodiments of this application;

[0022] Figure 2 This is a flowchart of the redundancy configuration method for the reactor array in the embodiments of this application;

[0023] Figure 3 This is a first flowchart for determining a health score in an embodiment of this application;

[0024] Figure 4 This is a second flowchart for determining a health score in an embodiment of this application;

[0025] Figure 5 This is a flowchart illustrating the determination of the second failure probability in an embodiment of this application;

[0026] Figure 6 This is a schematic diagram of a computer device in an embodiment of this application. Detailed Implementation

[0027] To enable those skilled in the art to better understand the present application, the technical solutions in specific embodiments of the present application will be clearly and completely described below with reference to the accompanying drawings.

[0028] It should be noted that the terms "first," "second," and similar terms used in this application specification and claims do not indicate any order, quantity, or importance, but are merely used to distinguish different components. Similarly, "an" or "a" and similar terms do not indicate a quantity limitation, but rather indicate the presence of at least one. "A plurality" or "several" indicates at least two. "Comprising" or "including" and similar terms mean that the elements or objects preceding "comprising" or "including" encompass the elements or objects listed following "comprising" or "including" and their equivalents, and do not exclude other elements or objects. "Connected" or "linked" and similar terms are not limited to physical or mechanical connections, but can include electrical connections, whether direct or indirect.

[0029] This application provides a method for redundant configuration of a reactor array 300. In one embodiment, this method for redundant configuration of the reactor array 300 can be applied to, for example... Figure 1 In the redundancy configuration device 100 of the reactor array shown, the redundancy configuration method of the reactor array 300 includes a data acquisition module 11, a data processing module 12, and an execution module 13. The data acquisition module 11 is used to determine the health score of each reactor 31 in the reactor array 300 before producing the target sample. The data processing module 12 can perform some of the method steps in the following embodiments based on the health scores determined by the data acquisition module 11. The execution module 13 can activate several spare reactors to replace the faulty reactor 31.

[0030] In another embodiment, the redundancy configuration method can also be applied to a computer device, which can be a terminal device or a server. It should be noted that the computer device is not limited to the terminal devices and servers mentioned above; any device with data processing capabilities can be used as the computer device in this embodiment. The computer device is communicatively connected to the redundancy configuration device 100. Exemplarily, the communication connection can be a wired connection using a data bus as the transmission medium or a wireless connection. The redundancy configuration device 100 can determine the health score of each reactor 31 in the reactor array 300 before producing the target sample and send the health score to the computer device, causing the computer device to execute the method steps described in the following embodiments.

[0031] In another embodiment, the redundancy configuration method is also applied to the redundancy configuration device 100 and the computer device, wherein some steps of the redundancy configuration method are performed by the redundancy configuration device 100 and some steps are performed by the computer device.

[0032] like Figure 2 As shown, this application provides a method for redundant configuration of a reactor array 300, which includes the following steps:

[0033] Step S201: Determine the health score of each reactor 31 in the reactor array 300 before producing the target sample.

[0034] Among them, the health score can reflect whether any abnormality has occurred in each reactor 31.

[0035] In some implementations, if the health score If so, the reactor 31 is judged to be in a healthy state; if If so, the reactor 31 is judged to be in a sub-healthy state; if If so, the reactor 31 is determined to be in a fault state.

[0036] Among them, a healthy state indicates that the reactor 31 is normal; a sub-healthy state indicates that the reactor 31 has an abnormality but can still maintain operation; and a faulty state indicates that the reactor 31 cannot maintain normal operation.

[0037] like Figure 3 As shown, in some implementations, determining the health score of each reactor 31 in the reactor array 300 before producing the target sample specifically includes the following steps:

[0038] Step S301: Sample the process data of the reactor array 300 during the production of the target sample.

[0039] The process data includes temperature, pressure, current, stirring speed, and material volume.

[0040] In some implementations, the number of data acquisition modules 11 is not less than the number of reactors 31 in the reactor array 300. Each reactor 31 is equipped with a data acquisition module 11 to collect process data from each reactor 31. Some data acquisition modules 11 may also be located in the transmission channel between two reactors 31. For a data acquisition module 11 in a reactor 31, it may include at least one of a temperature sensor, a pressure sensor, a current sensor, a speed sensor, and a weighing sensor. Among them, the temperature sensor is used to detect the temperature inside the reactor 31, and the temperature sensor includes, but is not limited to, thermocouple sensors and resistance temperature detectors (RTDs); the pressure sensor is used to detect the pressure inside the reactor 31, and the pressure sensor includes, but is not limited to, diffused silicon pressure transmitters and explosion-proof pressure transmitters; the current sensor is used to detect the current inside the reactor 31, and the current sensor includes, but is not limited to, Hall effect current sensors and coil current sensors; the speed sensor is used to detect the stirring speed inside the reactor 31, and the speed sensor includes, but is not limited to, Hall effect speed sensors and electrical induction speed sensors; the weighing sensor is used to detect the material capacity inside the reactor 31 by weight conversion, and the weighing sensor includes, but is not limited to, shear beam weighing sensors and column weighing sensors.

[0041] Step S302: Extract the first time series feature dataset from the process data, and input the first time series feature dataset into the preset time series neural network model to obtain the second time series feature dataset after dimensionality reduction.

[0042] The first time series feature dataset is used to indicate information such as temperature, pressure, current, stirring speed, and material capacity in the reactor 31.

[0043] The pre-defined Temporal Convolutional Network (TCN) model can reduce the dimensionality of the first time series feature dataset through dilated causal convolution and residual connections. While reducing the data size, it removes redundant information and obtains a second time series feature dataset that is concise, structurally normalized, and has clear features, which is convenient for subsequent calculation and analysis.

[0044] Step S303: Based on the historical production dataset of the production target sample, and combined with the second time series feature dataset, determine the target feature vector dataset for the current moment.

[0045] The historical production dataset for the target sample consists of digitized records of the "best quality batch" from the historical production history of the target sample. The data types within this dataset are no less than those in the process data (temperature, pressure, current, stirring speed, material volume, etc.). The "best quality batch" is the optimal target sample selected from the historical production records according to preset criteria, including but not limited to the target sample exhibiting the most stable performance and / or the highest corrosion resistance. The target feature vector dataset indicates the residual between the historical production dataset and the current process data, i.e., it indicates the difference between the current target sample and the "best quality batch" in the historical production records.

[0046] Step S304: Based on the weighted Mahalanobis distance of the target feature vector dataset, determine the health score of each reactor 31 in the reactor array 300.

[0047] like Figure 4 As shown, based on the weighted Mahalanobis distance of the target feature vector dataset, the health score of each reactor 31 in the reactor array 300 is determined, specifically including the following steps:

[0048] Step S401: Obtain the inverse matrix of the covariance matrix of the reference feature dataset based on the historical production dataset, and generate the attention weight matrix corresponding to the reaction process at the current moment based on the process map.

[0049] The process map is built based on historical production datasets, and the attention weight matrix is ​​a d×d diagonal weight matrix, which is used to indicate the weight of process data in different reaction processes and assign the weight to the corresponding features.

[0050] Step S402: Calculate the weighted statistical distance value based on the target feature vector dataset, the inverse matrix, and the attention weight matrix.

[0051] In some implementations, the weighted statistical distance values ​​satisfy the following relationship:

[0052] ;

[0053] In the formula, This represents the weighted statistical distance value. This represents the transpose of the target feature vector dataset. Represents the attention weight matrix. Represents the inverse matrix. This represents the target feature vector dataset.

[0054] Step S403: Obtain the health score corresponding to the weighted statistical distance value based on the preset mapping relationship.

[0055] In some implementations, the health score satisfies the following relationship:

[0056]

[0057] In the formula, This represents the health score, where i represents the i-th reactor (31). This represents the normal operating threshold of reactor 31, i.e., the allowable deviation range when reactor 31 is operating normally, which is obtained from several weighted statistical distance values.

[0058] Step S202: Determine the first failure probability of each reactor based on the differences in the health scores of each reactor using the distribution function.

[0059] The distribution functions include, but are not limited to, the Weibull distribution, the exponential distribution, and the gamma distribution.

[0060] It should be noted that different reactors 31 have different states, and the probability of first failure of each reactor 31 is different due to the state differences. The state differences include at least one of the following: the number of times the reactor 31 is used, the lifespan of the parts, and the replacement time of the parts. Moreover, the probability of first failure will change in real time after the reactor array 300 is put into production.

[0061] In some implementations, the first failure probability satisfies the following relationship:

[0062]

[0063] In the formula, Indicates the probability of the first failure. Indicates the change over time. Indicates shape parameters, Indicates characteristic lifetime.

[0064] Specifically, the shape parameters are determined by the physical properties of the reactor 31, and they can represent the relationship between the first failure probability and the change over time. This indicates that the probability of the first failure decreases as the amount of change over time increases. This indicates that the probability of the first failure is constant. This indicates that the probability of the first failure increases as the amount of change over time increases.

[0065] Characteristic lifespan is related to health score For example, the characteristic lifespan is positively correlated with the health score, i.e., the characteristic lifespan satisfies... As the health score decreases, the characteristic lifespan also decreases, and the probability of the first failure of reactor 31 increases.

[0066] It should be noted that, in determining the first failure probability of each reactor 31, in addition to using the Weibull distribution, mapping methods including but not limited to the exponential distribution and the Gamma distribution can also be used.

[0067] In some implementations, the data processing module 12 is connected to each data acquisition module 11 to obtain the health score of each reactor 31 determined by each data acquisition module 11.

[0068] Through the above steps, based on the health score of each reactor 31, the real-time characteristic life of each reactor 31 is obtained, and then the dynamic first failure probability of each reactor 31 is obtained, thereby improving the accuracy and real-time performance of determining the first failure probability of each reactor 31.

[0069] Step S203: Based on the Poisson binomial distribution of the first failure probability of each reactor 31, determine the second failure probability of multiple reactors 31 in the reactor array 300 failing simultaneously.

[0070] like Figure 5 As shown, in some implementations, the second failure probability of multiple reactors 31 in the reactor array 300 simultaneously failing is determined based on the Poisson binomial distribution of the first failure probability of each reactor 31. This specifically includes the following steps:

[0071] Step S501: Calculate the distribution of the first failure probability using the Poisson binomial distribution model.

[0072] Step S502: Solve using a numerical algorithm based on recursion or discrete Fourier transform to obtain the second failure probability of multiple reactors 31 in the reactor array 300 failing simultaneously.

[0073] In some implementations, the second failure probability satisfies the following relationship:

[0074]

[0075] in, This indicates the number of 31 reactors that failed simultaneously within a single batch cycle. It is a non-negative integer (e.g., 0, 1, 2, …). Let K represent the probability that the random variable takes the value k, and A represent a specific subset of reaction vessels. This represents the set of all subsets of reactors of size k. This indicates that the combination of 31 reactors 31 that may contain k faulty reactors is traversed, and i indicates that it belongs to a subset. In a certain reactor 31, j indicates that it belongs to One of the reaction vessels 31, This indicates that the reactor 31 is not malfunctioning.

[0076] For example, the reactor array 300 includes 100 reactors 31, and F2 is the set of all subsets that contain exactly 2 reactors 31.

[0077] In some implementations, the data processing module 12 can determine a second failure probability of simultaneous failure of multiple reactors 31 in the reactor array 300 based on a first failure probability.

[0078] Through the above steps, the second failure probability of k reactors 31 failing simultaneously is determined.

[0079] Step S204: Substitute the second failure probability into the pre-constructed total cost function to obtain the target value when the function value of the total cost function is minimized.

[0080] The target value is used to indicate the number of failures in reactor 31. The total cost function includes a risk cost function, which indicates the maximum expected loss caused by the scrapping of the target sample when the number of failures exceeds the total number of spare reactors in reactor array 300.

[0081] In some implementations, the risk-cost function The following relationship must be satisfied:

[0082] ;

[0083] In the formula, α Indicates the confidence interval. S Indicates the total number of spare reactors. C batch Let P represent the value of the material and P represent the second failure probability. K=k This indicates the number of malfunctions in reactor 31, where M represents the total number of reactors 31 currently in operation. The risk aversion coefficient represents the degree of impact of a failure of reactor 31 on the utility of reactor array 300.

[0084] It should be noted that the maximum expected loss is positively correlated with the material value of the production target sample.

[0085] For example, the total number of spare vessels S The value range is [0, 10], and the total number of spare reactors is... S It is an integer.

[0086] In some implementations, the total cost function also includes a holding cost function and a switching loss function. The holding cost function indicates the power-on preheating cost and depreciation cost of activating the standby reactor; the switching loss function indicates the material cleaning loss of the malfunctioning reactor 31.

[0087] In some implementations, the total cost function The following relationship must be satisfied:

[0088]

[0089] In the formula, This represents the power-on preheating cost and depreciation cost of a spare reactor; E represents the material cleaning loss of one reactor 31; E represents the expected cost. This represents the expected number of switching operations, which is calculated from the first failure probability of each reactor 31.

[0090] In this implementation, by traversing... In order to obtain Find the minimum value and determine it. Number of spare reactors when taking the minimum value , That is the target value.

[0091] In some implementations, the data processing module 12 can substitute the second failure probability into a pre-constructed total cost function to obtain the target value when the function value of the total cost function is minimized.

[0092] Step S205: According to the target value, wake up the corresponding number of spare reactors and perform a preheating operation to replace the faulty reactor 31.

[0093] In some implementations, the execution module 13 can acquire the target value and, based on the target value, activate a corresponding number of backup reactors to replace the faulty reactor 31.

[0094] The redundant configuration method for the reactor array 300 provided in this application determines the health score of each reactor 31 in the reactor array 300 before producing the target sample. Based on the health score, it determines the first failure probability of each reactor 31 and the second failure probability of multiple reactors 31 failing simultaneously. The second failure probability is then incorporated into a pre-constructed total cost function. Based on various costs when switching to a backup reactor, a target value with the lowest total cost is obtained. The corresponding number of backup reactors are then activated and preheated according to the target value, allowing the backup reactors to replace the failed reactor 31. After the reactor array 300 is put into production, the number of backup reactors activated is dynamically adjusted based on the individual differences of each reactor 31, thereby reducing property losses when a reactor 31 fails.

[0095] like Figure 6 As shown, this application also provides a computer device 200, which includes a memory 21 and a processor 22. The memory 21 stores a computer program. When the computer program is executed by the processor 22, the processor 22 performs the steps of the above-described method for redundant configuration of the reactor array.

[0096] Specifically, processor 22 may include a central processing unit, or an application-specific integrated circuit (ASIC), or one or more integrated circuits that can be configured to implement embodiments of the present invention.

[0097] In some implementations, memory 21 may include a large-capacity memory for data or instructions.

[0098] For example, the memory includes a hard disk drive (HDD), a floppy disk drive, flash memory, an optical disk, a magneto-optical disk, a universal serial bus (USB) drive, or any combination of the above-mentioned memory.

[0099] For example, memory 21 may be located inside or outside the computer device 200.

[0100] In some possible implementations, the computer device 200 also includes a communication interface 23 and a bus 24. The processor 22, memory 21, and communication interface 23 are connected via the bus 24 and communicate with each other.

[0101] The communication interface 23 is mainly used to realize communication between various modules, devices, units and / or equipment in the embodiments of this application.

[0102] Bus 24 includes hardware and / or software that couples components of computer device 200 together. For example, and not to limit, bus 24 may include an accelerated graphics port or other graphics bus, an enhanced industry standard architecture bus, a front-side bus, a low pin count bus, a memory bus, or other suitable bus or any combination of the above buses.

[0103] This application also provides a computer-readable storage medium (not shown) storing a computer program that, when executed by a processor, implements the above-described method for redundant configuration of the reactor array.

[0104] The computer-readable storage medium includes, but is not limited to, electronic, magnetic, optical, infrared, or other physical storage devices or equipment, which may contain or store information such as executable instructions, data, etc. More specific examples of computer-readable storage media include electrical connections based on one or more wires, RAM (Random Access Memory), volatile memory, non-volatile memory, flash memory, storage drives (such as hard disk drives), SSDs (Solid State Disks), any type of storage disk (such as optical discs), or similar memory, or any suitable combination of the foregoing. This application also provides a computer-readable storage medium (not shown) storing a computer program that, when executed, implements the aforementioned redundant configuration method for a reactor array.

[0105] Computer-readable storage media include, but are not limited to, electronic, magnetic, optical, infrared, or other physical storage devices or apparatuses that may contain or store information such as executable instructions, data, etc. More specific examples of computer-readable storage media include electrical connections based on one or more wires, RAM (Random Access Memory), volatile memory, non-volatile memory, flash memory, storage drives (such as hard disk drives), SSDs (Solid State Disks), any type of storage disk (such as optical discs), or similar memory, or any suitable combination of the foregoing.

[0106] It should be understood that those skilled in the art can make improvements or modifications based on the above description, and all such improvements and modifications should fall within the protection scope of the appended claims.

Claims

1. A method for redundant configuration of a reactor array, characterized in that, Determine the health score of each reactor in the reactor array before producing the target sample, wherein the health score is determined by a weighted Mahalanobis distance; The first failure probability of each reactor is determined by using a distribution function based on the differences in the health scores of each reactor; wherein the first failure probability of each reactor varies due to state differences, and the state differences include at least one of the following: differences in the number of times the reactor is used, differences in the lifespan of parts, and differences in the replacement time of parts; The distribution of the first failure probability is calculated using a Poisson binomial distribution model; The numerical algorithm based on recursion or discrete Fourier transform is used to solve the problem and obtain the second failure probability of multiple reactors in the reactor array failing simultaneously. The second failure probability is substituted into the pre-constructed total cost function to obtain the target value when the function value of the total cost function is minimized. The target value is used to indicate the number of failures of the reactor. The total cost function includes a risk cost function, which is used to indicate the maximum expected loss caused by the number of failures exceeding the total number of spare reactors in the reactor array, resulting in the scrapping of the target sample. The corresponding number of the backup reactors are activated according to the target value to replace the malfunctioning reactors; The determination of the health score for each reactor based on the weighted Mahalanobis distance includes: The process data of the reactor array during the production of the target sample are sampled; Extract the first time series feature dataset from the process data, and input the first time series feature dataset into a preset time-series neural network model to obtain the second time series feature dataset after dimensionality reduction; Based on the historical production dataset of the target sample, the target feature vector dataset at the current moment is determined by combining the second time series feature dataset. Based on the historical production dataset, the inverse matrix of the covariance matrix of the reference feature dataset is obtained, and an attention weight matrix corresponding to the reaction process at the current moment is generated based on the process map. The weighted statistical distance value is calculated based on the target feature vector dataset, the inverse matrix, and the attention weight matrix. The health score is obtained based on a preset mapping relationship and the weighted statistical distance value.

2. The redundancy configuration method according to claim 1, characterized in that, The total cost function also includes a holding cost function and a switching loss function; The holding cost function is used to indicate the power-on preheating cost and depreciation cost of waking up the standby vessel; The switching loss function is used to indicate the material cleaning loss of the reactor that malfunctions.

3. The redundancy configuration method according to claim 2, characterized in that, The total cost function The following relationship must be satisfied: ； In the formula, This represents the power-on preheating cost and depreciation cost of one of the aforementioned spare reactors. E represents the material washing loss of one of the aforementioned reactors; E represents the expected cost. Indicates the expected number of switching operations.

4. The redundancy configuration method according to claim 1, characterized in that, The maximum expected loss is positively correlated with the material value of producing the target sample.

5. The redundancy configuration method according to claim 1, characterized in that, The risk cost function The following relationship must be satisfied: ； In the formula, α Indicates the confidence interval. S This indicates the total number of spare reactors. C batch Let P represent the value of the material and the second failure probability. K=k M represents the number of malfunctions in the reactor, and M represents the total number of reactors currently in operation. This represents the risk aversion coefficient.

6. The redundancy configuration method according to claim 1, characterized in that, The determination of the health score of each reactor in the reactor array before producing the target sample includes: The process data of the reactor array during the production of the target sample are sampled; Extract the first time series feature dataset from the process data, and input the first time series feature dataset into a preset time-series neural network model to obtain the second time series feature dataset after dimensionality reduction; Based on the historical production dataset of the target sample, the target feature vector dataset at the current moment is determined by combining the second time series feature dataset. The target feature vector dataset is used to indicate the residual between the historical production dataset and the process data at the current moment. The health score of each reactor in the reactor array is determined based on the weighted Mahalanobis distance of the target feature vector dataset.

7. A redundant configuration device for a reactor array, characterized in that, include: The data acquisition module is used to determine the health score of each reactor in the reactor array before producing the target sample, wherein the health score is determined by the weighted Mahalanobis distance; The data processing module is used to determine a first failure probability of each reactor based on the differences in the health scores of each reactor using a distribution function; wherein the first failure probability of each reactor varies due to state differences, and the state differences include at least one of the following: differences in the number of times the reactor is used, differences in the lifespan of parts, and differences in the replacement time of parts; based on the Poisson binomial distribution of the first failure probability of each reactor, a second failure probability of multiple reactors in the reactor array failing simultaneously is determined; the second failure probability is substituted into a pre-constructed total cost function to obtain a target value that minimizes the function value of the total cost function, and the target value is used to indicate the number of reactor failures; wherein the total cost function includes a risk cost function, and the risk cost function is used to indicate the maximum expected loss caused by the number of failures exceeding the total number of spare reactors in the reactor array, resulting in the scrapping of the target sample; An execution module is used to wake up a corresponding number of the backup reactors according to the target value, so as to replace the malfunctioning reactors; The determination of the health score for each reactor based on the weighted Mahalanobis distance includes: The process data of the reactor array during the production of the target sample are sampled; Extract the first time series feature dataset from the process data, and input the first time series feature dataset into a preset time-series neural network model to obtain the second time series feature dataset after dimensionality reduction; Based on the historical production dataset of the target sample, the target feature vector dataset at the current moment is determined by combining the second time series feature dataset. Based on the historical production dataset, the inverse matrix of the covariance matrix of the reference feature dataset is obtained, and an attention weight matrix corresponding to the reaction process at the current moment is generated based on the process map. The weighted statistical distance value is calculated based on the target feature vector dataset, the inverse matrix, and the attention weight matrix. The health score is obtained based on a preset mapping relationship and the weighted statistical distance value.

8. A computer device, characterized in that, It includes a memory and a processor, wherein the memory stores a computer program that, when executed by the processor, performs the redundancy configuration method according to any one of claims 1-6.

9. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores program instructions that, when executed on one or more processors, implement the redundancy configuration method according to any one of claims 1-6.

Images