Large-flow high-precision nitrogen-oxygen-gas mixed flow system

Abstract

The invention relates to the technical field of nitrogen and oxygen mixed flow, and discloses a large-flow high-precision nitrogen and oxygen mixed flow system. The system comprises a gas attribute feature library construction module for periodically collecting historical component data to generate a feature library containing flow and concentration reference values; the real-time gas component monitoring module is used for acquiring the real-time component concentration and flow value of the current nitrogen and oxygen; the mixed flow deviation analysis module is used for generating a component deviation matrix according to the reference value and the real-time value; the dynamic parameter optimization module retrieves the matched flow fluctuation data set, and when the dispersion exceeds a threshold value, the proportion of the total duration from the historical start to the current moment is calculated to serve as the gas compensation proportion; and the mixed flow execution control module is used for adjusting input proportioning parameters of nitrogen and oxygen from a historical starting moment based on the compensation proportion. According to the system, all the modules work cooperatively, and high-precision mixing of nitrogen and oxygen under large flow is achieved.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of nitrogen-oxygen gas mixing, in particular to a large-flow high-precision nitrogen-oxygen gas mixing system. BACKGROUND

[0002] In industrial production, scientific experiments and other scenarios, it is often necessary to use nitrogen-oxygen mixed gas with a specific ratio. Different application scenarios have different requirements for the flow and component concentration of the mixed gas, especially in some fields sensitive to gas parameters, the precision of the mixed gas directly affects the quality of the final product or the reliability of the experimental results.

[0003] Currently, existing nitrogen-oxygen gas mixing systems often face many challenges when dealing with large-flow gas mixing. Because gas itself has fluidity and diffusivity, it is easily affected by external environmental factors such as temperature and pressure changes during transportation. These factors can cause fluctuations in gas flow and component concentration, affecting mixing precision. Traditional mixing control methods mostly use fixed ratio parameters for regulation and lack dynamic response capability to real-time data. When gas flow or component concentration deviates, it cannot be adjusted in time, making it difficult to ensure the quality of the mixed gas.

[0004] Existing systems also have deficiencies in data collection and analysis. Most systems can only monitor current gas parameters, ignoring the accumulation and application of historical data. Without a complete gas attribute feature library, it is difficult to achieve high-precision mixing control when analyzing deviation causes and optimizing parameters due to the lack of effective data support. At the same time, there is a lack of scientific and reasonable algorithm support in the logic of deviation analysis and parameter optimization, making the adjusted ratio parameters often unable to adapt well to changes in actual working conditions, further affecting the performance of the mixing system. SUMMARY

[0005] The purpose of the present application is to provide a large-flow high-precision nitrogen-oxygen gas mixing system to solve the problems raised in the background art.

[0006] To achieve the above purpose, the present application provides a large-flow high-precision nitrogen-oxygen gas mixing system, which comprises: A gas attribute feature library construction module configured to periodically collect historical component data sets of the target mixed gas, and generate a feature library containing gas flow reference values and component concentration reference values; A real-time gas component monitoring module configured to obtain real-time component concentrations and real-time flow values of nitrogen and oxygen in the current conveying pipeline; The mixed flow deviation analysis module is configured to generate a component deviation matrix according to the flow reference value and the component concentration reference value of the gas attribute feature library construction module and the real-time flow value and the real-time component concentration of the real-time gas component monitoring module; The dynamic parameter optimization module is configured to retrieve a gas flow fluctuation data set matching the component deviation matrix, and calculate a total time length proportion from a historical collection start time to a current time as a gas compensation proportion when a dispersion of the gas flow fluctuation data set exceeds a preset dispersion threshold. The mixed flow execution control module is configured to adjust input proportioning parameters of nitrogen and oxygen starting from the historical collection start time based on the gas compensation proportion.

[0007] Preferably, the gas attribute feature library construction module comprises: The reference parameter acquisition unit is configured to acquire a preset component concentration constraint condition of the target mixed gas. The fluctuation analysis unit is configured to take the preset component concentration constraint condition as a boundary and statistically analyze a periodic fluctuation range of gas flow according to a historical sample database. The feature library updating unit is configured to continuously update the periodic fluctuation range of gas flow until all constraint conditions are met when a new component concentration constraint condition is added, and set 0.6 times of the minimum fluctuation range as a gas collection period.

[0008] Preferably, the gas attribute feature library construction module further comprises: The storage capacity monitoring unit is configured to monitor a storage space occupation threshold of gas component historical data. The feature library triggering unit is configured to start collection of a historical component data set of the target mixed gas when a total amount of the gas component historical data reaches the storage space occupation threshold.

[0009] Preferably, the fluctuation analysis unit comprises: The component data retrieval subunit is configured to extract a plurality of flow record values and corresponding time stamps meeting the preset component concentration constraint condition from the historical sample database when detecting a gas flow parameter. The flow deviation calculation subunit is configured to calculate a difference between each two of the plurality of flow record values to generate a flow record deviation modulus set and a time stamp deviation set. The feature screening subunit is configured to extract a flow record selected modulus set greater than a preset time deviation threshold from the time stamp deviation set. The main distribution identification subunit is configured to perform main distribution density analysis on the flow record selected modulus set to output the periodic fluctuation range of the gas flow.

[0010] Preferably, the fluctuation analysis unit comprises: A concentration mapper calling subunit configured to call a concentration mapping processor bound with a gas component type and the preset component concentration constraint when detecting a gas component concentration parameter; An association parameter extraction subunit configured to retrieve, from the historical sample database, a plurality of sets of input flow characteristic values and corresponding component concentration record values that meet the preset component concentration constraint; A concentration mapping conversion subunit configured to convert the input flow characteristic values by the concentration mapping processor to generate a set of component concentration characteristic values; A fluctuation range generation subunit configured to generate a periodic fluctuation range of the gas flow according to the set of component concentration characteristic values and the corresponding component concentration record values.

[0011] Preferably, the system further comprises: A component type configuration module configured to set a concentration mapping relationship of a target gas component through a user terminal; A mapping processor training module configured to collect a plurality of sets of training data to train the concentration mapping processor with preset gas equipment models and component concentration constraints as boundaries; A processor binding module configured to establish a binding relationship between the concentration mapping processor and the target gas component and the preset component concentration constraint.

[0012] Preferably, the dynamic parameter optimization module comprises: A matrix decomposition unit configured to separate the component deviation matrix into a flow deviation vector and a component concentration deviation vector; A cloud retrieval unit configured to send an encrypted retrieval request to the cloud, upload the gas type, equipment model, flow deviation vector, and component concentration deviation vector after encryption using a key, and receive a gas flow fluctuation data set returned by the cloud; An abnormal data processing unit configured to perform an outlier filtering operation on the gas flow fluctuation data set.

[0013] Preferably, the cloud retrieval unit comprises: A cloud decryption subunit configured to parse the encrypted data by a cloud decryption module, extract the gas type, equipment model, flow deviation vector, and component concentration deviation vector to construct a retrieval constraint condition; A node communication subunit configured to initiate a retrieval request to a distributed node through the cloud and obtain a node verification key; A node retrieval subunit configured to encrypt the retrieval constraint condition using the node verification key, generate a node retrieval instruction and send it to the distributed node, and receive the returned gas flow fluctuation data set.

[0014] Preferably, the mixed flow execution control module comprises: a valve opening calculation unit configured to generate opening adjustment amounts of nitrogen input valves and oxygen input valves according to the gas compensation ratio; a real-time ratio feedback unit configured to transmit valve opening change data to the real-time gas component monitoring module in real time to trigger new real-time flow value collection; a closed-loop control unit configured to iteratively adjust the input ratio parameters based on updated real-time flow values until the component deviation matrix is zeroed.

[0015] Preferably, the closed-loop control unit comprises: a historical data comparison subunit configured to call the historical component data set of the gas attribute feature library construction module to verify the consistency of the current input ratio parameters and the historical optimal ratio; a deviation convergence judgment subunit configured to send a termination optimization instruction to the dynamic parameter optimization module when the modulus value change amount of the component deviation matrix in three consecutive iterations is less than a convergence threshold; a ratio parameter solidification subunit configured to write the final input ratio parameters into the storage unit of the gas attribute feature library construction module.

[0016] Compared with the prior art, the present application has the following beneficial effects: Through the gas attribute feature library construction module, the historical component data set of the target mixed gas is periodically collected to generate a feature library containing gas flow reference values and component concentration reference values, providing rich historical data references for subsequent operation of the system. These historical data can reflect the laws and characteristics of gas mixing in different time periods, so that the system can have comparable references when processing real-time data, thereby more accurately judging whether the current gas mixing state meets the expectations.

[0017] The real-time gas component monitoring module can timely obtain the real-time component concentration and real-time flow value of nitrogen and oxygen in the current delivery pipeline, so that the system can real-time master the dynamic situation of gas mixing. This real-time monitoring capability enables the system to quickly discover changes in gas parameters, which saves time for subsequent deviation analysis and parameter adjustment, and avoids the decline in mixing accuracy caused by information lag.

[0018] The mixed flow deviation analysis module generates a component deviation matrix according to the reference values in the feature library and the real-time values obtained by the real-time monitoring module. This matrix form can clearly show the deviation degree and specific situation between the current gas component and the reference values. The generation of the deviation matrix makes the originally complex deviation information become organized and visualized, which helps the system to more accurately locate the deviation, and provides a clear direction for subsequent parameter optimization.

[0019] The dynamic parameter optimization module retrieves the gas flow fluctuation data set matching the component deviation matrix, and calculates the proportion of the total time length from the historical acquisition starting time to the current time as the gas compensation ratio when the dispersion of the data set exceeds the preset threshold. This compensation ratio calculation method based on historical data and real-time deviation fully considers the time factor and fluctuation in the gas mixing process, making the determination of the compensation ratio more reasonable and better adapting to the dynamic changes of the gas flow.

[0020] The mixed flow execution control module adjusts the input ratio parameters of nitrogen and oxygen based on the gas compensation ratio from the historical acquisition starting time. This adjustment method can trace back to the starting point of the historical data and comprehensively adjust the historical trend and real-time deviation. In this way, the system can gradually correct the deviation in the gas mixing process, making the mixing ratio of nitrogen and oxygen closer to the expected value, thereby realizing high-precision gas mixing under large flow. BRIEF DESCRIPTION OF DRAWINGS

[0021] Figure 1 The working principle diagram of the large-flow high-precision nitrogen-oxygen gas mixed flow system according to the present application; Figure 2 The flowchart of the gas attribute feature library construction module; Figure 3 The flowchart of the flow analysis of the fluctuation analysis unit; Figure 4 The flowchart of the concentration analysis of the fluctuation analysis unit; Figure 5 The flowchart of the data processing of the dynamic parameter optimization module. DETAILED DESCRIPTION

[0022] The technical solutions in the embodiments of the present application will be clearly and completely described below with reference to the drawings in the embodiments of the present application. Obviously, the described embodiments are only part of the embodiments of the present application, not all. Based on the embodiments in the present application, all other embodiments obtained by those skilled in the art without creative labor are within the scope of protection of the present application.

[0023] Please refer to Figure 1 The present application provides a large-flow high-precision nitrogen-oxygen gas mixed flow system, which comprises a gas attribute feature library construction module, a real-time gas component monitoring module, a mixed flow deviation analysis module, a dynamic parameter optimization module and a mixed flow execution control module. The specific implementation is as follows: The gas attribute feature library construction module periodically acquires a set of historical component data of the target mixed gas, and generates a feature library containing gas flow reference values and component concentration reference values.

[0024] The real-time gas component monitoring module continuously acquires the real-time component concentration and real-time flow value of nitrogen and oxygen in the current conveying pipeline.

[0025] The mixed flow deviation analysis module calls the flow reference value and component concentration reference value in the gas attribute feature library construction module, combines the real-time flow value and real-time component concentration obtained by the real-time gas component monitoring module, and generates a component deviation matrix through matrix operation.

[0026] The dynamic parameter optimization module retrieves a gas flow fluctuation data set matched with the component deviation matrix, and when the dispersion of the data set exceeds a preset dispersion threshold, calculates the total duration ratio from the historical acquisition starting time to the current time, and takes the ratio as the gas compensation ratio.

[0027] The mixed flow execution control module adjusts the input ratio parameters of nitrogen and oxygen from the historical acquisition starting time based on the above gas compensation ratio, to realize accurate mixing of nitrogen and oxygen gas.

[0028] Embodiment 1: refer to Figure 2 The reference parameter acquisition unit in the gas attribute feature library construction module is responsible for receiving the preset component concentration constraint conditions of the target mixed gas. These constraint conditions are determined based on the actual application scenarios of the mixed gas, and the constraint conditions in different scenarios are different. For example, in the medical oxygen scene, the oxygen concentration in the mixed gas may need to be controlled at 93%±3%, and the nitrogen concentration is adjusted accordingly; while in some industrial welding scenarios, the oxygen concentration may be required to be between 20%-25%, and the nitrogen concentration is maintained at 75%-80%. These specific concentration ranges are the preset component concentration constraint conditions, which are stored by the reference parameter acquisition unit as the basis data for subsequent analysis.

[0029] The fluctuation analysis unit processes the data in the historical sample database with the preset component concentration constraint conditions as the boundary. The historical sample database stores the flow data of nitrogen and oxygen, the corresponding component concentration data and the time information when the system runs in different time periods in the past. The fluctuation analysis unit will filter out the historical data that meets the preset constraint conditions, and then statistically analyze the gas flow parameters in these data. Specifically, the data will be divided by time period, such as by hour, day or week, to observe the flow changes in each period. By comparing the maximum and minimum values of the flow in different periods, the fluctuation range of the flow under the condition of meeting the concentration constraint condition is determined. For example, in a certain period, the oxygen flow may fluctuate between 50-60L / min, and the nitrogen flow may fluctuate between 200-220L / min, which will be recorded as the preliminary periodic fluctuation range.

[0030] The role of the feature library updating unit is reflected when there are new component concentration constraints. New constraints may come from changes in application scenarios, such as originally only requiring nitrogen and oxygen concentrations, and now adding a constraint that the impurity content in the mixed gas must not exceed 0.001%. At this time, the feature library updating unit will re-call the historical sample database, filter data that meets both the original and new constraints, and re-statistical the periodic fluctuation range of gas flow. This process may need to be iterated several times, because the new constraint may exclude some of the data that originally met the conditions, causing the fluctuation range to change. Until the statistical fluctuation range can meet all the constraints, the update process will stop.

[0031] After determining the minimum fluctuation range that meets all the constraints, the feature library updating unit will set 0.6 times of the minimum fluctuation range as the gas collection period. For example, assuming that after analysis, the time span corresponding to the minimum fluctuation range is 10 minutes, then the gas collection period is set to 6 minutes. The reason for this setting is that collecting data with a period smaller than the minimum fluctuation range can capture more detailed changes in flow, ensuring that the collected data can fully reflect the periodic fluctuation characteristics of the flow, providing more accurate basic data for subsequent deviation analysis and parameter optimization. At the same time, the feature library updating unit will also update the newly determined flow reference value, component concentration reference value, and collection period, etc. information to the gas property feature library, ensuring that the data in the feature library always matches the current constraints, supporting the continuous and stable operation of the system.

[0032] During the operation of the feature library updating unit, it will monitor in real time whether there are new constraint conditions input. Once a new constraint condition is detected, the above update process will be started immediately without human intervention, ensuring that the system can automatically adapt to changes in application scenarios. At the same time, during the update of the fluctuation range, historical fluctuation range data will be preserved for use when backtracking analysis is needed, such as comparing the differences in flow fluctuation under different constraints.

[0033] Embodiment 2: see Figure 3The gas attribute feature library construction module comprises a storage capacity monitoring unit and a feature library triggering unit, and the fluctuation analysis unit is composed of a component data retrieval subunit, a flow deviation calculation subunit, a feature screening subunit, and a main distribution identification subunit. The storage capacity monitoring unit continuously monitors the storage space occupied by the gas component historical data. The system will pre-set a storage space occupation threshold. The setting of the threshold is related to the storage medium capacity of the system, the data preservation period, and the data acquisition frequency. For example, if the system uses a 100 GB storage medium and needs to save 3 months of historical data, the storage space occupation threshold can be set to 80 GB to reserve a certain amount of redundant space to prevent storage overflow. The storage capacity monitoring unit calculates the occupation ratio of the current gas component historical data by reading the used space and total capacity of the storage medium in real time. When the ratio reaches the pre-set threshold, a signal will be sent to the feature library triggering unit.

[0034] After receiving the signal from the storage capacity monitoring unit, the feature library triggering unit starts the collection of the historical component data set of the target mixed gas. The collection process covers the flow data, component concentration data, and corresponding time information of nitrogen and oxygen. These data will be stored in the historical sample database for subsequent fluctuation analysis.

[0035] When the fluctuation analysis unit is running, the component data retrieval subunit plays a role first. When the system detects the gas flow parameters, this subunit will filter out multiple sets of flow record values and corresponding time stamps that meet the conditions from the historical sample database according to the pre-set component concentration constraints. For example, if the pre-set oxygen concentration needs to be within 21% ± 1%, the component data retrieval subunit will extract all oxygen flow values and corresponding time information recorded within this concentration range, as well as nitrogen flow values and time stamps at the same period, ensuring the time correlation of the data.

[0036] The flow deviation calculation subunit processes the extracted multiple sets of flow record values and calculates the flow deviation modulus in a pairwise comparison manner. Specifically, for any two sets of flow record values, the absolute value of their difference is calculated to form a set of flow record deviation modulus; at the same time, the difference between the time stamps corresponding to the two sets of records is calculated to form a set of time stamp deviation. For example, there are two sets of oxygen flow record values, 50 L / min and 52 L / min, corresponding to time stamps t1 and t2, respectively. The flow deviation modulus is 2 L / min, and the time stamp deviation is t2-t1.

[0037] The feature screening subunit screens the selected modulus set of flow records from the timestamp deviation set that is greater than the preset time deviation threshold. The preset time deviation threshold is set according to the preliminary judgment of the system on the flow fluctuation period, for example, set to 30 minutes, so that the modulus of the flow record deviation whose timestamp deviation exceeds 30 minutes is retained, and the modulus of the deviation in a short period of time is excluded to avoid the influence of short-period disturbance on the analysis result.

[0038] The main distribution identification subunit performs main distribution density analysis on the selected modulus set. By counting the frequency of the modulus, a probability density curve is drawn to determine the main interval of the modulus distribution. For example, if most of the modulus values are concentrated in the range of 1-3 L / min, this interval is determined as the main distribution interval of the flow deviation, and the corresponding flow range is the periodic fluctuation range of the gas flow.

[0039] Each unit works cooperatively to form a complete data analysis chain from the triggering of data acquisition to the determination of the flow fluctuation range, ensuring that historical data can effectively support the precise control of nitrogen-oxygen mixed gas by the system. As the historical sample database continues to accumulate, the results of the fluctuation analysis will gradually be optimized, making the determined periodic fluctuation range more in line with the actual operation situation.

[0040] Embodiment 3: refer to Figure 4 The fluctuation analysis unit includes a concentration mapper calling subunit, an associated parameter extraction subunit, a concentration mapping conversion subunit, and a fluctuation range generation subunit. The system also has a component type configuration module, a mapping processor training module, and a processor binding module. The component type configuration module provides a visual operation interface, through which the user can set the concentration mapping relationship of the target gas component, such as defining the mapping curve type between the nitrogen flow and the nitrogen concentration, or setting the corresponding relationship between the oxygen flow rate and the oxygen concentration rate. These mapping relationships can be adjusted according to different gas mixing scenarios to adapt to diversified concentration control needs.

[0041] The mapping processor training module carries out training work with preset gas equipment models and component concentration constraints as boundaries. The preset gas equipment model determines the adaptation range of the processor, for example, for a specific model of gas mixing valve and flow meter, the training data should come from the actual operation records of the equipment of this model. The component concentration constraint limits the selection range of the training data, only the operation data that meets the concentration requirement will be included in the training set. During the training process, multiple sets of input flow characteristic values and corresponding component concentration values are collected as training data, the input flow characteristic values include instantaneous flow, flow peak value, flow average value, etc. The parameters of the mapping processor are continuously optimized through iterative calculation, so that the processor can accurately output the corresponding component concentration characteristics according to the input flow characteristic values.

[0042] The processor binding module binds the trained concentration mapping processor with the target gas component and the preset component concentration constraint condition, and stores the binding relationship in the mapping relationship library of the system. When a specific gas component needs to be processed or a specific concentration constraint condition is met, the system can quickly call the corresponding concentration mapping processor, thereby improving the data processing efficiency.

[0043] When detecting the gas component concentration parameter, the concentration mapper calling subunit calls the bound concentration mapping processor from the mapping relationship library according to the current gas component type and the preset component concentration constraint condition. For example, when detecting the oxygen concentration, the mapping processor bound with oxygen and the current concentration constraint condition is called.

[0044] The correlation parameter extraction subunit retrieves a plurality of sets of input flow characteristic values and corresponding component concentration record values that meet the preset component concentration constraint condition from the historical sample database. During the retrieval process, the time consistency of the input flow characteristic values and the component concentration record values is ensured through timestamp matching, the data of the two sets is ensured to be derived from the same running period, and the correlation and reliability of the data are ensured.

[0045] The concentration mapping conversion subunit converts the input flow characteristic values through the called concentration mapping processor to generate a set of component concentration characteristic values. During the conversion process, the processor maps the flow characteristic values to the corresponding concentration characteristic values according to the mapping relationship obtained through training, for example, converts the oxygen flow average value in a time period to the oxygen concentration average value in the time period, forms a one-to-one corresponding characteristic value pair, and the like.

[0046] The fluctuation range generation subunit calculates the deviation value between the set of component concentration characteristic values and the corresponding component concentration record values, that is, the difference between each component concentration characteristic value and the corresponding record value. By statistically analyzing the distribution range of these deviation values, the maximum value and the minimum value of the deviation values are determined, and the corresponding flow range is the periodic fluctuation range of the gas flow. For example, if the deviation values are concentrated between-0.5% and 0.5%, the corresponding flow change range is the periodic fluctuation range of the oxygen flow.

[0047] In the concentration mapping conversion process, the component concentration characteristic value is calculated by using the following formula:

[0048] wherein, the component concentration characteristic value is represented by C, the mapping coefficient is represented by M, the input flow characteristic value is represented by F, and The correction coefficient is represented. Both the mapping coefficient and the correction coefficient are trained by the mapping processor training module according to historical data, and reflect the linear relationship between the flow characteristic value and the concentration characteristic value. Through the formula, the input flow characteristic value can be quantitatively converted into the component concentration characteristic value.

[0049] The fluctuation range generation subunit feeds back the determined periodic fluctuation range of the gas flow to the gas attribute feature library construction module, so as to update the flow reference value in the feature library, and enable the feature library to accurately reflect the fluctuation characteristics of the gas flow.

[0050] Embodiment 4: refer to Figure 5 The dynamic parameter optimization module includes a matrix decomposition unit, a cloud retrieval unit, and an abnormal data processing unit. The cloud retrieval unit is further divided into a cloud decryption subunit, a node communication subunit, and a node retrieval subunit.

[0051] The matrix decomposition unit receives the component deviation matrix generated by the mixed flow deviation analysis module. The matrix is a two-dimensional array calculated based on real-time monitoring data and reference values in the feature library. The row vector represents deviation records at different time points, and the column vector distinguishes the deviation parameters of nitrogen and oxygen. The matrix decomposition unit decomposes the component deviation matrix into a flow deviation vector and a component concentration deviation vector through an eigenvalue decomposition algorithm. The flow deviation vector is composed of eigenvalues related to flow in the matrix, and the component concentration deviation vector corresponds to eigenvalues related to concentration. The two vectors respectively reflect the deviation degrees of flow and concentration. For example, after decomposition of a certain component deviation matrix, the flow deviation vector may be [0.3, -0.2, 0.1], indicating that the flow at the three time points deviates from the reference value by +0.3, -0.2, and +0.1 units, respectively; and the component concentration deviation vector is [-0.1, 0.4, -0.2], corresponding to the deviation of the concentration.

[0052] When the cloud retrieval unit is started, it first performs encryption processing on the gas type, the equipment model, the flow deviation vector, and the component concentration deviation vector. Encryption uses a system-preset asymmetric encryption algorithm. The private key is stored in the local encryption module, and the public key is used for data encryption to ensure that information is not leaked during transmission. For example, the "nitrogen-oxygen mixed system", "MIX-2000 model", and the above-mentioned deviation vectors are encrypted into a specific string and uploaded to the cloud.

[0053] After receiving the encrypted data, the cloud decryption subunit calls the cloud decryption module, uses the paired private key to decrypt, extracts the original information, and constructs the retrieval constraint condition. The retrieval constraint condition is presented in the form of key-value pairs, such as "gas type: nitrogen-oxygen mixture", "equipment model: MIX-2000", "flow deviation range: [-0.3, 0.4]", etc., which is used to limit the retrieval range of the distributed nodes.

[0054] The node communication subunit sends a retrieval request to the distributed node cluster through the cloud, and each node stores gas flow fluctuation data of a specific region or time period. During communication, the node returns its node verification key, which is a one-time temporary key used only for encryption of this retrieval instruction to enhance communication security.

[0055] The node retrieval subunit encrypts the retrieval constraint conditions again using the node verification key to generate a node retrieval instruction. For example, the constraint conditions are converted into binary instructions and encrypted to ensure that the node can only analyze retrieval requests related to itself. After receiving the instruction, the distributed node matches the gas flow fluctuation data set that meets the conditions from the local database. The data set contains records of flow changes under similar deviation conditions in history, such as a node returning a data set that may contain 50 flow fluctuation curves, each curve corresponding to a different deviation repair process.

[0056] The abnormal data processing unit cleans the received gas flow fluctuation data set and uses the box plot method to identify outliers: first calculate the quartiles Q1 (lower quartile) and Q3 (upper quartile) of the data set, and determine the reasonable range as [Q1-1.5×IQR, Q3+1.5×IQR] (IQR=Q3-Q1). Data outside this range is marked as an outlier and removed. For example, a data set has Q1=2.1, Q3=3.5, and IQR=1.4, and the reasonable range is [0.0, 5.6]. The value of 6.2 is determined to be abnormal and deleted.

[0057] The following is a gas flow fluctuation data set (part) returned by a node before and after abnormal processing:

[0058] The processed data set will be used for dispersion calculation in the dynamic parameter optimization module. The entire cloud retrieval process realizes safe and efficient calling of cross-regional data through hierarchical encryption and distributed node cooperation, and the abnormal processing guarantees the data quality.

[0059] Example 5: The mixed flow execution control module is composed of a valve opening calculation unit, a real-time ratio feedback unit, and a closed-loop control unit, wherein the closed-loop control unit includes a historical data comparison subunit, a deviation convergence judgment subunit, and a ratio parameter solidification subunit.

[0060] The valve opening degree calculation unit receives the gas compensation ratio output by the dynamic parameter optimization module, which is an adjustment coefficient calculated based on the historical time length proportion. The calculation unit internally stores the characteristic curve of the nitrogen input valve and the oxygen input valve, which is obtained through previous calibration tests and reflects the corresponding relationship between the valve opening degree and the actual gas flow. For example, the characteristic curve of a certain oxygen valve shows that for every 10% increase in opening degree, the flow increases by 5 L / min. The valve opening degree calculation unit reverses the opening degree change amount that needs to be adjusted according to the gas compensation ratio and the characteristic curve. If the compensation ratio requires a 10% increase in oxygen flow, combining the characteristic curve can calculate that the valve opening degree needs to increase by 20%, and according to the nitrogen-oxygen ratio, the opening adjustment amount of the nitrogen valve is calculated to ensure that the adjustments of the two are matched with each other.

[0061] The real-time ratio feedback unit converts the calculated valve opening adjustment amount into an electrical signal and sends it to the valve actuator to drive the valve to adjust the opening degree. At the same time, the unit sends real-time data of the valve opening degree change to the real-time gas component monitoring module through the data transmission interface, triggering the module to start a new round of data collection. After receiving the trigger signal, the real-time gas component monitoring module controls the sensors in the pipeline to continuously collect real-time flow values and component concentration values three times within a preset time interval, taking the average of the three measurements as the feedback data after this adjustment to reduce the influence of single measurement error.

[0062] The closed-loop control unit starts the iterative adjustment process after receiving the new real-time flow value. The historical data comparison sub-unit retrieves the historical component data set from the gas attribute characteristic library construction module, selects historical operation records similar to the current working condition, and extracts the optimal input ratio parameters in these records. For example, historical records with an environmental temperature and pipeline pressure deviation within ±5% of the current value are selected, and the ratio parameters that make the mixed gas components most stable in these records are taken as the reference benchmark, which is compared with the current input ratio parameters item by item to analyze the differences and possible reasons.

[0063] The deviation convergence judgment sub-unit continuously monitors the component deviation matrix generated by the mixed flow deviation analysis module and calculates the change amount of the matrix module value in each iteration process. The matrix module value is obtained by taking the square root of the sum of the squares of all elements in the matrix, reflecting the size of the overall deviation. When the module value change amount calculated by three consecutive iterations is less than the preset convergence threshold, it indicates that the current input ratio parameter has approached a stable state, and further adjustment has limited effect on improving the mixing effect. At this time, the deviation convergence judgment sub-unit sends a termination optimization instruction to the dynamic parameter optimization module to suspend the parameter optimization process to avoid system fluctuations caused by excessive adjustment.

[0064] The matching parameter solidification subunit formats the current input matching parameter after receiving the termination optimization instruction, converts it into a record conforming to the data format of the gas attribute feature library, including parameter value, effective time, corresponding working condition label and other information. Subsequently, the record is written into the storage unit of the gas attribute feature library construction module through the feature library interface, and an index is established to facilitate subsequent retrieval. For example, the parameter is stored in association with the "high flow rate-normal temperature" working condition label. When the system encounters the same working condition again, the parameter can be directly called as the initial setting, shortening the adjustment time.

[0065] The entire closed-loop control process repeatedly repeats the cycle of valve adjustment, data feedback, and parameter optimization until all elements in the component deviation matrix generated by the mixed flow deviation analysis module are zero, i.e., the real-time component concentration and flow rate value are completely consistent with the reference value. At this time, the mixed gas reaches the preset accuracy requirement, and the mixed flow execution control module stops the adjustment action, maintaining the current valve opening and input matching parameter.

[0066] It should be noted that, in this document, relational terms such as first and second and the like can be used solely to distinguish one entity or action from another entity or action without necessarily requiring or implying any actual such relationship or order between such entities or actions. Moreover, the terms "comprises", "comprising", or any other variations thereof, are intended to cover a non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements does not include only those elements but can include other elements not expressly listed or inherent to such process, method, article, or apparatus.

[0067] Although embodiments of the present application have been shown and described, it is to be understood that various modifications, substitutions, replacements and variations can be made to these embodiments without departing from the principles and spirit of the present application, and the scope of the present application is defined by the appended claims and their equivalents.

Claims

1. A high-flow high-precision nitrogen-oxygen gas mixing system, characterized in that, The method comprises the following steps: A gas attribute feature library construction module is configured to periodically collect a set of historical component data of a target mixed gas, and generate a feature library containing gas flow reference values and component concentration reference values; A real-time gas component monitoring module is configured to obtain real-time component concentration and real-time flow value of nitrogen and oxygen in a current conveying pipeline; A mixed flow deviation analysis module is configured to generate a component deviation matrix according to the flow reference values and the component concentration reference values of the gas attribute feature library construction module, and the real-time flow value and the real-time component concentration of the real-time gas component monitoring module; A dynamic parameter optimization module is configured to retrieve a gas flow fluctuation data set matching the component deviation matrix, and calculate a total time length proportion from a historical collection starting time to a current time as a gas compensation ratio when a dispersion degree of the gas flow fluctuation data set exceeds a preset dispersion threshold; A mixed flow execution control module is configured to adjust input proportioning parameters of nitrogen and oxygen from the historical collection starting time based on the gas compensation ratio.

2. The high-precision nitrogen-oxygen gas mixing system of claim 1, wherein, The gas attribute feature library construction module comprises: A reference parameter acquisition unit is configured to obtain a preset component concentration constraint condition of the target mixed gas; A fluctuation analysis unit is configured to take the preset component concentration constraint condition as a boundary, and statistically analyze a periodic fluctuation range of gas flow according to a historical sample database; A feature library updating unit is configured to continuously update the periodic fluctuation range of the gas flow until all constraint conditions are met when a new component concentration constraint condition is added, and set 0.6 times of the minimum fluctuation range as a gas collection period.

3. The high-precision nitrogen-oxygen gas mixing system of claim 1, wherein, The gas attribute feature library construction module further comprises: A storage capacity monitoring unit is configured to monitor a storage space occupation threshold of gas component historical data; A feature library triggering unit is configured to start collection of a set of historical component data of the target mixed gas when a total amount of the gas component historical data reaches the storage space occupation threshold.

4. The high-precision nitrogen-oxygen gas mixing system of claim 2, wherein, The fluctuation analysis unit comprises: A component data retrieval subunit is configured to extract a plurality of flow record values and corresponding time stamps that meet the preset component concentration constraint condition from the historical sample database when detecting a gas flow parameter; A flow deviation calculation subunit is configured to calculate a difference between each two of the plurality of flow record values to generate a flow record deviation modulus set and a time stamp deviation set; A feature screening subunit is configured to extract a flow record selected modulus set greater than a preset time deviation threshold from the time stamp deviation set; A main distribution identification subunit is configured to perform main distribution density analysis on the flow record selected modulus set, and output the periodic fluctuation range of the gas flow.

5. The high-precision nitrogen-oxygen gas mixing system of claim 2, wherein, The fluctuation analysis unit comprises: A concentration mapper calling subunit is configured to call a concentration mapping processor bound with a gas component type and the preset component concentration constraint condition when detecting a gas component concentration parameter; An associated parameter extraction subunit is configured to retrieve a plurality of input flow feature values and corresponding component concentration record values that meet the preset component concentration constraint condition from the historical sample database; The concentration mapping conversion subunit is configured to convert the input flow characteristic value by the concentration mapping processor to generate a set of component concentration characteristic values; The fluctuation range generation subunit is configured to generate a periodic fluctuation range of the gas flow according to the set of component concentration characteristic values and the corresponding component concentration record value.

6. The high-precision nitrogen-oxygen gas mixing system of claim 5, wherein, Further comprising: The component type configuration module is configured to set a concentration mapping relationship of a target gas component through a user terminal; The mapping processor training module is configured to collect multiple sets of training data to train the concentration mapping processor with preset gas equipment models and component concentration constraints as boundaries; The processor binding module is configured to establish a binding relationship between the concentration mapping processor and the target gas component and the preset component concentration constraint.

7. The high-precision nitrogen-oxygen gas mixing system of claim 1, wherein, The dynamic parameter optimization module includes: The matrix decomposition unit is configured to separate the component deviation matrix into a flow deviation vector and a component concentration deviation vector; The cloud retrieval unit is configured to send an encrypted retrieval request to the cloud, upload the gas type, equipment model, flow deviation vector, and component concentration deviation vector after encryption using a key, and receive a gas flow fluctuation data set returned by the cloud; The abnormal data processing unit is configured to perform an outlier filtering operation on the gas flow fluctuation data set.

8. The high-precision nitrogen-oxygen gas mixing system of claim 7, wherein, The cloud retrieval unit includes: The cloud decryption subunit is configured to parse the encrypted data through a cloud decryption module, extract the gas type, equipment model, flow deviation vector, and component concentration deviation vector to construct a retrieval constraint condition; The node communication subunit is configured to initiate a retrieval request to a distributed node through the cloud and obtain a node verification key; The node retrieval subunit is configured to encrypt the retrieval constraint condition using the node verification key, generate a node retrieval instruction and send it to the distributed node, and receive the returned gas flow fluctuation data set.

9. The high-precision nitrogen-oxygen gas mixing system of claim 1, wherein, The mixed flow execution control module includes: The valve opening calculation unit is configured to generate opening adjustment amounts of a nitrogen input valve and an oxygen input valve according to the gas compensation ratio; The real-time ratio feedback unit is configured to transmit valve opening change data to the real-time gas component monitoring module in real time to trigger new real-time flow value collection; The closed-loop control unit is configured to iteratively adjust the input ratio parameters based on updated real-time flow values until the component deviation matrix is zero.

10. The high-precision nitrogen-oxygen gas mixing system of claim 9, wherein, The closed-loop control unit includes: The historical data comparison subunit is configured to call a historical component data set of the gas attribute feature library construction module to verify the consistency of the current input ratio parameters and the historical optimal ratio; The deviation convergence judgment subunit is configured to send a termination optimization instruction to the dynamic parameter optimization module when the modulus value change of the component deviation matrix in three consecutive iterations is less than a convergence threshold; The ratio parameter solidification subunit is configured to write the final input ratio parameters to the storage unit of the gas attribute feature library construction module.

Images