A method, system, device, and medium for monitoring the operational state of a mass spectrometer
By constructing a vacuum effect analysis network and a mass axis offset analysis network in the mass spectrometer, and combining environmental parameters for status monitoring, the problem of untimely fault warning in the existing technology is solved, and earlier and more accurate fault identification is achieved.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- HEFEI PUJIA MEDICAL LAB CO LTD
- Filing Date
- 2025-06-24
- Publication Date
- 2026-06-05
AI Technical Summary
Existing methods for monitoring the operational status of mass spectrometers cannot capture the coupling relationships between parameters, resulting in untimely fault warnings.
By collecting the operating environment parameters of the mass spectrometer, a vacuum influence analysis network and a mass axis offset analysis network are constructed using machine learning to obtain vacuum degree influence parameters and mass axis offset parameters. Combined with the vacuum verification coefficient and the mass verification coefficient, anomaly analysis of the fusion state is performed to identify and warn of faults.
It improves the timeliness of mass spectrometer fault warnings, avoids the distortion of warning results caused by single-dimensional prediction bias, and enables accurate fault identification at an earlier stage.
Smart Images
Figure CN120687987B_ABST
Abstract
Description
Technical Field
[0001] This invention relates to the field of fault monitoring, and in particular to a method, system, device and medium for monitoring the operating status of a mass spectrometer. Background Technology
[0002] In existing technologies, the monitoring of the operating status of mass spectrometers is often limited to the threshold judgment of a single parameter. However, this method can only determine whether a single parameter exceeds the preset threshold, but it cannot capture the coupling relationship between parameters and it is difficult to discover hidden influencing factors. As a result, the fault warning is not timely during the monitoring of the operating status of the mass spectrometer. Summary of the Invention
[0003] This invention addresses the technical problem of untimely fault warnings during the monitoring of mass spectrometer operating status in the prior art, and provides a method, system, equipment, and medium for monitoring the operating status of a mass spectrometer.
[0004] The technical solution of the present invention to solve the above-mentioned technical problems is as follows:
[0005] In a first aspect, the present invention provides a method for monitoring the operating status of a mass spectrometer, comprising:
[0006] Environmental parameters in the operating environment of the mass spectrometer are collected, and the influence of the vacuum degree of the mass spectrometer is analyzed based on the environmental parameters to obtain the vacuum degree influence parameters.
[0007] Based on the environmental parameters, an analysis of the impact of mass spectrometer mass axis offset was performed to obtain mass axis offset parameters.
[0008] The actual vacuum degree influence parameters and actual mass error parameters of the mass spectrometer are obtained through testing. The state influence is verified by combining the vacuum degree influence parameters and the mass axis offset parameters. The vacuum verification coefficient and mass verification coefficient are then obtained through processing.
[0009] Based on the vacuum verification coefficient and the mass verification coefficient, a fusion state anomaly analysis is performed on the vacuum degree influence parameter and the mass axis offset parameter to obtain the state anomaly coefficient, and a fault identification and early warning is performed.
[0010] Secondly, the present invention provides a mass spectrometer operating status monitoring system, comprising:
[0011] The vacuum analysis module is used to collect environmental parameters in the operating environment of the mass spectrometer, and to perform vacuum degree influence analysis based on the environmental parameters to obtain vacuum degree influence parameters.
[0012] The axis offset analysis module is used to analyze the impact of the mass spectrometer mass axis offset based on the environmental parameters, and to obtain the mass axis offset parameters.
[0013] The state verification module is used to test and obtain the actual vacuum degree influence parameters and actual mass error parameters of the mass spectrometer, and to perform state influence verification by combining the vacuum degree influence parameters and mass axis offset parameters, and to process and obtain vacuum verification coefficient and mass verification coefficient.
[0014] The anomaly warning module is used to perform a fusion state anomaly analysis on the vacuum degree influence parameter and the mass axis offset parameter based on the vacuum verification coefficient and the mass verification coefficient, obtain the state anomaly coefficient, and perform fault identification and warning.
[0015] Thirdly, the present invention provides an electronic device, comprising:
[0016] Memory, used to store computer software programs;
[0017] The processor is used to read and execute the computer software program, thereby realizing the mass spectrometer operation status monitoring method provided in this application.
[0018] Fourthly, the present invention provides a computer-readable storage medium storing a computer program, which, when executed by a processor, implements a method for monitoring the operating status of a mass spectrometer as described in the first aspect.
[0019] The beneficial effects of this invention are:
[0020] Compared to existing technologies, this application first collects environmental parameters from the mass spectrometer's operating environment. Based on these parameters, it analyzes the impact of vacuum degree on the mass spectrometer, obtaining vacuum degree impact parameters. This allows for the assessment of the mass spectrometer's operating status, providing a reliable basis for monitoring its operational status. Secondly, it analyzes the impact of mass axis offset on the mass spectrometer based on the environmental parameters, obtaining mass axis offset parameters, which provide a reliable basis for subsequent results analysis. Thirdly, it tests and obtains the actual vacuum degree impact parameters and actual mass error parameters of the mass spectrometer. Combining these parameters with the mass axis offset parameters, it verifies the state impact, processing to obtain vacuum verification coefficients and mass verification coefficients. The actual data validates the model's predicted data and provides data support for subsequent fault identification and early warning. Finally, based on the vacuum verification coefficient and the mass verification coefficient, a fusion state anomaly analysis is performed on the vacuum degree influence parameter and the mass axis offset parameter to obtain the state anomaly coefficient, and a fault identification and early warning is performed. The basic state anomaly coefficient is compensated and amplified by the absolute difference between the vacuum verification coefficient and the mass verification coefficient, avoiding the distortion of the early warning result caused by the single-dimensional prediction bias. This allows for more accurate fault identification and early warning at an earlier stage, so that relevant technical personnel can carry out operation and maintenance.
[0021] Through the above technical solution, this application uses a vacuum influence analysis network and a mass axis offset analysis network constructed based on machine learning. By taking environmental parameters as input, it outputs vacuum degree influence parameters and mass axis offset parameters, and compensates and expands the basic state anomaly coefficient accordingly. In this way, fault identification and early warning can be made at an earlier stage, improving the timeliness of mass spectrometer fault warning. Attached Figure Description
[0022] Figure 1 A flowchart illustrating a method for monitoring the operating status of a mass spectrometer provided by the present invention;
[0023] Figure 2 A schematic diagram of a mass spectrometer operation status monitoring system provided by the present invention;
[0024] Figure 3 This is a schematic diagram of the structure of the electronic device provided by the present invention.
[0025] In the attached diagram, the components represented by each number are as follows:
[0026] Vacuum analysis module 11, shaft offset analysis module 12, status verification module 13, anomaly warning module 14, electronic equipment 200, memory 210, processor 220, computer program 211. Detailed Implementation
[0027] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.
[0028] In the description of this invention, the terms "first" and "second" are used for descriptive purposes only and should not be construed as indicating or implying relative importance or implicitly specifying the number of indicated technical features. Thus, a feature defined as "first" or "second" may explicitly or implicitly include one or more of the stated features. In the description of this invention, "a plurality of" means two or more, unless otherwise explicitly specified.
[0029] In the description of this invention, the term "for example" is used to mean "used as an example, illustration, or description." Any embodiment described as "for example" in this invention is not necessarily to be construed as being more preferred or advantageous than other embodiments. The following description is provided to enable any person skilled in the art to make and use the invention. Details are set forth in the following description for purposes of explanation. It should be understood that those skilled in the art will recognize that the invention can be made without using these specific details. In other instances, well-known structures and processes will not be described in detail to avoid obscuring the description of the invention with unnecessary detail. Therefore, the invention is not intended to be limited to the embodiments shown, but is consistent with the broadest scope of the principles and features disclosed herein.
[0030] Example 1, as Figure 1 As shown, this embodiment of the invention provides a method for monitoring the operating status of a mass spectrometer, including:
[0031] S10: Collect environmental parameters in the operating environment of the mass spectrometer, and perform vacuum degree influence analysis based on the environmental parameters to obtain vacuum degree influence parameters;
[0032] Vacuum level is one of the key performance indicators of a mass spectrometer, and it is a fundamental condition for ensuring the stable operation of the instrument. Specifically, vacuum level refers to the degree of vacuum environment inside the mass spectrometer. The fewer the number of gas molecules and the lower the gas pressure, the higher the vacuum level. That is, vacuum level and pressure have an inverse relationship: the lower the pressure, the higher the vacuum level, and vice versa. To ensure the quality of mass spectrometry operation, it is typically necessary to maintain a vacuum level of 10... -3 ~10 -6 In a high vacuum state, if the pressure is too high, ions will frequently collide with gas molecules, significantly reducing ion transport efficiency and even preventing them from reaching the detector. Furthermore, a high vacuum environment effectively reduces ion source filament oxidation and mass analyzer contamination, thus extending the lifespan of the instrument hardware. Further, temperature is a crucial factor affecting the vacuum level of a mass spectrometer. Increased temperature leads to thermal expansion and contraction, causing small gaps in the vacuum seal, deteriorating the high vacuum state, and increasing internal pressure. For example, from 10... -3 Pa becomes 10 -2 Pa, which in turn leads to a decrease in vacuum level, and once the vacuum level decreases, the probability of ions colliding with gas molecules during transport increases significantly, affecting the normal detection of the mass spectrometer. Therefore, the trend of vacuum level change can be predicted by parameters such as ambient temperature, and the operating status of the mass spectrometer can be monitored accordingly.
[0033] To address the aforementioned issues, this application collects environmental parameters (including temperature) from the operating environment of the mass spectrometer and analyzes the impact of these environmental parameters on the vacuum level of the mass spectrometer to obtain parameters affecting the vacuum level.
[0034] Specifically, step S10 in the method includes:
[0035] Collect environmental parameters in the operating environment of the mass spectrometer, including ambient temperature;
[0036] The environmental parameters are input into the vacuum influence analysis network of the mass spectrometer, and the vacuum influence parameters are identified and output.
[0037] In this embodiment, environmental parameters in the operating environment of the mass spectrometer are first collected. These parameters include ambient temperature, as temperature is a crucial factor affecting the vacuum level of the mass spectrometer; increased temperature leads to a decrease in vacuum. For example, the ambient temperature with a timestamp can be collected in real time using a temperature sensor deployed in the mass spectrometer's operating environment. For instance, the collected ambient temperature might be 25°C (16:00 on June 5, 2025).
[0038] Secondly, the environmental parameters are input into the vacuum influence analysis network of the mass spectrometer, and the vacuum influence parameters are identified and output. Specifically, the vacuum influence analysis network is built based on machine learning. Inputting environmental parameters (such as ambient temperature) and processing them through multiple layers of neurons, it outputs vacuum influence parameters (such as vacuum change parameters). The larger the vacuum influence parameter, the greater the deviation from the ideal state, indicating a worse current operating state of the mass spectrometer. For example, inputting environmental parameters (such as ambient temperature: 25℃) into the vacuum influence analysis network, the vacuum influence parameters (such as vacuum change parameter: 0.2) are identified and output.
[0039] Specifically, the training steps for the "vacuum influence analysis network" include:
[0040] Based on the operation and monitoring data of the mass spectrometer, a set of sample environmental parameters was collected, and the vacuum degree change parameters of the mass spectrometer under different sample environmental parameters were collected. The set of parameters affecting the sample vacuum degree was then labeled.
[0041] Construct a machine learning-based network for vacuum effect analysis;
[0042] Under the set of sample environmental parameters and the set of sample vacuum degree influence parameters, the vacuum influence analysis network is trained and the parameters are optimized under supervision. The training is completed after the accuracy is tested and converged.
[0043] In this embodiment, training data is first prepared. Specifically, based on the operational monitoring data of the mass spectrometer, a set of sample environmental parameters (e.g., ambient temperature: 25℃, 30℃, 35℃, 40℃) is collected, along with vacuum degree variation parameters of the mass spectrometer under different sample environmental parameters (e.g., 0.2, 0.4, 0.7, 0.9), and a set of parameters affecting the sample vacuum degree is obtained. Further, the set of parameters affecting the sample vacuum degree is divided into a training set, a validation set, and a test set according to a ratio of 7:1.5:1.5, which are used as training data for the vacuum influence analysis network.
[0044] The vacuum degree change parameter is calculated as follows: (Vacuum degree at current temperature - Reference vacuum degree) / Reference vacuum degree. This is because the ideal operating temperature for most mass spectrometers (such as quadrupole mass spectrometers and time-of-flight mass spectrometers) is 18℃ to 25℃. Temperatures below 15℃ can lead to decreased performance of circuit components, poor lubrication of mechanical parts, and instability of the vacuum system. Temperatures above 30℃ can cause a decrease in vacuum degree, an increase in the probability of ion collisions, and a decrease in sensitivity. Therefore, a reference temperature (e.g., 22℃) can be determined based on the ideal operating temperature of the mass spectrometer, and the vacuum degree corresponding to the reference temperature (e.g., 1×10⁻⁶) can be collected. -5 A reference vacuum level (Pa) is used as the baseline vacuum level, and the vacuum level variation parameters at different temperatures are analyzed accordingly. For example, the reference vacuum level corresponding to the reference temperature (e.g., 22°C) is (e.g., 1×10⁻⁶ Pa). -5 The vacuum degree of the mass spectrometer at 25℃ is 1.2 × 10 Pa. -5 Pa, then the vacuum degree change parameter at this time = (1.2 × 10 Pa) -5 Pa-1×10 -5 Pa) / 1×10 -5 With Pa=0.2, by collecting the vacuum level of the mass spectrometer at different temperatures, multiple vacuum level variation parameters can be calculated, which can then be used as a set of parameters affecting the sample vacuum level.
[0045] Secondly, a vacuum influence analysis network based on machine learning is constructed. This network is built using machine learning algorithms (such as LSTM and CNN), taking environmental parameters as input, processing them through multiple layers of neurons, and outputting vacuum influence parameters. It mainly consists of an input layer, hidden layers, and an output layer.
[0046] Specifically: The input layer serves as the interface between the vacuum influence analysis network and external data. It primarily receives the collected environmental parameters of the mass spectrometer (such as ambient temperature). These parameters are converted into numerical signals that the network can process and input to the next layer. The number of neurons in the input layer corresponds to the number of input parameters, with each neuron responsible for transmitting information about one parameter. For example, if only temperature is considered, the input layer contains one neuron. The hidden layer is the brain of the vacuum influence analysis network, primarily responsible for data feature extraction and complex relational calculations. It consists of multiple layers of neurons, with each layer's neurons connecting to the next layer via weighted connections. When processing environmental parameter data, the neurons in the hidden layer perform weighted summation of the input signals (such as ambient temperature), highlighting key features and introducing nonlinear factors through activation functions (such as ReLU and Sigmoid) to uncover the complex nonlinear relationship between environmental parameters and vacuum level changes. This application recommends setting 3-5 hidden layers, each containing 20 neurons. Those skilled in the art can adjust the number of layers and the number of neurons per layer according to actual data and task requirements. Generally speaking, the more layers and the more neurons, the stronger the network's expressive power, but it also increases computational complexity and the risk of overfitting. The output layer is mainly responsible for converting the feature information extracted and processed by the hidden layers into specific vacuum degree influencing parameters (such as the predicted vacuum pressure change value, vacuum degree decrease rate, etc.). The number of neurons in the output layer depends on the number of parameters that need to be output. For example, if only the vacuum pressure change value is output, then the output layer will only have 1 neuron.
[0047] Finally, model training is performed. Specifically, under the sample environmental parameter set and the sample vacuum degree influence parameter set, supervised training parameter optimization is performed on the vacuum influence analysis network, and training is completed after the test accuracy converges. For example, at the beginning of training, the weights and biases of the network are randomly initialized. The environmental parameters are input into the network through forward propagation, and the predicted value of vacuum degree influence is output after calculation by each layer of neurons. Subsequently, the difference between the predicted value and the true value is calculated using loss functions such as mean squared error (MSE). Then, the loss gradient is passed from the output layer to the input layer through the backpropagation algorithm, and the weights and biases are updated in combination with optimization algorithms such as Adam to reduce the loss. The training process adopts batch processing (e.g., Batch Size=32) and multi-epoch iteration strategy, and regularization (e.g., L2 regularization) and dropout techniques are used to prevent overfitting. When the loss value on the validation set no longer decreases for several consecutive epochs, or reaches a preset error threshold (e.g., 10), the training is completed. −4 If the model converges, training is stopped. Finally, the generalization ability of the model is evaluated using a test set to ensure that it can accurately predict the vacuum level change trend during the operation of the mass spectrometer, providing reliable support for equipment condition monitoring.
[0048] In summary, compared with existing technologies, this application collects environmental parameters (including temperature) in the operating environment of the mass spectrometer, analyzes the impact of these environmental parameters on the vacuum level of the mass spectrometer, obtains vacuum level impact parameters, and can use these parameters to assess the operating status of the mass spectrometer, providing a reliable basis for monitoring the operating status of the mass spectrometer.
[0049] S20: Analyze the impact of mass spectrometer mass axis offset based on the environmental parameters to obtain mass axis offset parameters;
[0050] The mass axis is one of the key performance indicators of a mass spectrometer, essentially representing the accuracy of measuring the mass-to-charge ratio of ions, and is commonly expressed as mass error (in ppm). In the qualitative analysis of compounds, especially when distinguishing substances with extremely small mass differences or isomers (e.g., C₂H₄O₂ and CH₄N₂O have very small theoretical mass differences), accurate differentiation is only possible if the mass error is controlled to <5 ppm. Therefore, a shift in the mass axis will lead to analytical errors. Furthermore, temperature is a significant factor affecting the mass axis shift of a mass spectrometer. Temperature causes thermal expansion and contraction of the magnetic material, resulting in a mass axis shift. Therefore, the mass axis shift parameter can be predicted using environmental parameters such as ambient temperature, and the operating status of the mass spectrometer can be monitored accordingly.
[0051] To address the aforementioned issues, this application analyzes the impact of mass spectrometer mass axis offset based on environmental parameters to obtain mass axis offset parameters.
[0052] Specifically, step S20 in the method includes:
[0053] The mass axis offset analysis network is invoked, wherein the mass axis offset analysis network is trained using a set of sample environment parameters and a set of sample mass axis offset parameters, and the sample mass axis offset parameters include the mass error magnitude;
[0054] The environmental parameters are input into the mass axis offset analysis network, and the mass axis offset parameters are obtained from the output.
[0055] In this embodiment, a pre-trained mass axis offset analysis network is first invoked. The structure and training steps of this network are similar to those of the vacuum influence analysis network in step S10. Specifically, the mass axis offset analysis network is constructed based on machine learning (such as LSTM, CNN, and other neural network algorithms). It takes environmental parameters as input, processes them through multiple layers of neurons, and outputs mass axis offset parameters. It mainly consists of an input layer, hidden layers, and an output layer. Further, the model is trained using a set of sample environmental parameters (e.g., ambient temperatures: 25℃, 30℃, 35℃, 40℃) and a set of sample mass axis offset parameters (e.g., mass error amplitudes: 0.25, 0.4, 0.7, 0.9) to obtain the mass axis offset analysis network.
[0056] Following the same logic as the preceding steps, the mass error amplitude = (mass error at current temperature - reference mass error) / reference mass error. A reference temperature (e.g., 22℃) can be determined based on the ideal operating temperature of the mass spectrometer, and the mass error corresponding to the reference temperature (e.g., 100ppm) is collected as the reference mass error. The mass error amplitude at different temperatures is then analyzed accordingly. For example, if the reference mass error corresponding to the reference temperature (e.g., 22℃) is (e.g., 100ppm), and the mass error of the mass spectrometer at 25℃ is 125ppm, then the mass error amplitude at this time = (125ppm - 100ppm) / 100ppm = 0.25. Thus, by collecting the mass error of the mass spectrometer at different temperatures, multiple mass error amplitudes can be calculated, which can then be used as a set of sample mass axis offset parameters.
[0057] Next, the environmental parameters are input into the mass axis offset analysis network, and the mass axis offset parameters are obtained from the output. For example, environmental parameters (such as ambient temperature: 28℃) are input into the mass axis offset analysis network, and the mass axis offset parameters (such as mass error amplitude: 0.3) are obtained from the output. The larger the mass axis offset parameter, the more it deviates from the ideal state, that is, the worse the current operating state of the mass spectrometer is.
[0058] In summary, compared with the prior art, this application analyzes the impact of mass axis offset on mass spectrometers based on environmental parameters, obtains mass axis offset parameters, and can use these parameters to assess the operating status of the mass spectrometer, providing a reliable basis for subsequent results analysis.
[0059] S30: Test and obtain the actual vacuum degree influence parameters and actual mass error parameters of the mass spectrometer, combine the vacuum degree influence parameters and mass axis offset parameters to verify the state influence, and process to obtain the vacuum verification coefficient and mass verification coefficient;
[0060] In the aforementioned steps, the vacuum degree influence parameters and quality error parameters predicted by the model may be affected by systematic biases (such as the error of the model algorithm itself) and random biases (such as accidental sensor noise), resulting in data deviation.
[0061] To address the aforementioned issues, this application obtains the actual vacuum level influence parameters and actual mass error parameters of the mass spectrometer through testing, and verifies the state influence by combining the vacuum level influence parameters and mass axis offset parameters, and processes them to obtain the vacuum verification coefficient and mass verification coefficient.
[0062] Specifically, step S30 in the method includes:
[0063] The actual vacuum level and actual mass error parameters of the mass spectrometer were obtained through testing.
[0064] Based on the vacuum degree influence parameters and the actual vacuum degree influence parameters, the vacuum state influence is verified to obtain the vacuum verification coefficient.
[0065] The impact of the quality status is verified based on the actual quality error parameters and the quality axis offset parameters to obtain the quality verification coefficient.
[0066] In this embodiment, the actual vacuum level influence parameters and actual mass error parameters of the mass spectrometer are first tested and obtained. For example, the actual vacuum level and actual mass error of the mass spectrometer are collected by sensors, and the actual vacuum level influence parameters and actual mass error parameters are calculated using the same method as in the preceding steps. These parameters reflect the current true operating state of the mass spectrometer.
[0067] Secondly, the vacuum state influence is verified based on the vacuum degree influence parameters and the actual vacuum degree influence parameters to obtain the vacuum verification coefficient. Specifically, the vacuum deviation amplitude between the vacuum degree influence parameters and the actual vacuum degree influence parameters is calculated, and then the vacuum verification coefficient is calculated based on the vacuum deviation amplitude. The vacuum verification coefficient reflects the degree of agreement between the predicted data and the actual detection data of the vacuum influence analysis network, indicating the prediction accuracy of the vacuum influence analysis network. The higher the prediction accuracy of the vacuum influence analysis network, the closer the vacuum degree influence parameters are to the actual vacuum degree influence parameters, the smaller the vacuum deviation amplitude, and the larger the vacuum verification coefficient.
[0068] Finally, the impact of the quality status is verified based on the actual quality error parameters and the quality axis offset parameters to obtain the quality verification coefficient. Specifically, similar to the calculation approach of the vacuum verification coefficient, the quality verification coefficient is obtained by calculating the magnitude of the quality error deviation between the quality axis offset parameters and the actual quality error parameters. The quality verification coefficient reflects the degree of agreement between the predicted data of the quality axis offset analysis network and the actual detection data, indicating the prediction accuracy of the quality axis offset analysis network. The higher the prediction accuracy of the quality axis offset analysis network, the closer the quality axis offset parameters and the actual quality error parameters are, the smaller the magnitude of the quality error deviation, and the larger the quality verification coefficient.
[0069] Furthermore, the step of "verifying the vacuum state influence based on the vacuum degree influence parameters and the actual vacuum degree influence parameters, and obtaining the vacuum verification coefficient" includes:
[0070] Calculate the vacuum deviation amplitude between the vacuum degree influence parameters and the actual vacuum degree influence parameters;
[0071] The vacuum verification coefficient is calculated based on the vacuum deviation amplitude.
[0072] In this embodiment, the vacuum deviation amplitude between the vacuum degree influence parameter and the actual vacuum degree influence parameter is first calculated. The vacuum deviation amplitude is calculated as: |vacuum degree influence parameter - actual vacuum degree influence parameter| / actual vacuum degree influence parameter. The vacuum deviation amplitude reflects the deviation between the predicted data and the actual data from the vacuum influence analysis network. The closer the vacuum degree influence parameter and the actual vacuum degree influence parameter are, the smaller the vacuum deviation amplitude. For example, if the vacuum influence analysis network predicts a vacuum degree influence parameter of 0.2 at 25°C (ambient temperature), and the actual vacuum degree at 25°C is collected by a sensor, the calculated actual vacuum degree influence parameter is 0.18, then the vacuum deviation amplitude is: |0.2 - 0.18| / 0.18 = 0.11.
[0073] Secondly, based on the vacuum deviation amplitude, a vacuum verification coefficient is calculated, where the vacuum verification coefficient = 1 - vacuum deviation amplitude. The vacuum verification coefficient reflects the degree of agreement between the predicted data and the actual data from the vacuum influence analysis network, i.e., it reflects the prediction accuracy of the vacuum influence analysis network. The higher the prediction accuracy of the vacuum influence analysis network, the closer the vacuum degree influence parameters are to the actual vacuum degree influence parameters, the smaller the vacuum deviation amplitude, and the larger the vacuum verification coefficient. For example, if the vacuum deviation amplitude is 0.11, then the vacuum verification coefficient = 1 - 0.11 = 0.89.
[0074] In summary, compared to existing technologies, this application obtains the actual vacuum level influence parameters and actual mass error parameters of the mass spectrometer through testing, and verifies the state influence by combining the vacuum level influence parameters and mass axis offset parameters, thus obtaining vacuum verification coefficients and mass verification coefficients. In this way, the model prediction data is verified through actual data, and data support is provided for subsequent fault identification and early warning.
[0075] S40: Based on the vacuum verification coefficient and the mass verification coefficient, perform a fusion state anomaly analysis on the vacuum degree influence parameter and the mass axis offset parameter to obtain the state anomaly coefficient and perform fault identification and early warning.
[0076] The aforementioned steps, based on environmental parameters (temperature), predict and output vacuum influence parameters and mass axis offset parameters through a vacuum influence analysis network and a mass axis offset analysis network. These two parameters can be fused to identify and warn of abnormal conditions in the mass spectrometer. However, in reality, both the vacuum influence analysis network and the mass axis offset analysis network may have prediction errors, and these errors may be asynchronous. For example, the vacuum influence analysis network may predict accurately, while the mass axis offset analysis network may predict inaccurately. Directly fusing these two parameters may cause the prediction deviation of the mass axis offset analysis network to be masked by the accurate prediction of the vacuum influence analysis network, resulting in distorted results and failing to trigger timely warnings.
[0077] To address the aforementioned issues, this application uses the vacuum verification coefficient and the mass axis offset coefficient to perform a fusion state anomaly analysis on the vacuum degree influence parameters and the mass axis offset parameters, obtains the state anomaly coefficient, and performs fault identification and early warning.
[0078] Specifically, step S40 in the method includes:
[0079] The verification error coefficient is calculated based on the vacuum verification coefficient and the mass verification coefficient.
[0080] Obtain the average vacuum level and average mass axis offset parameters when the mass spectrometer experiences abnormal conditions.
[0081] Calculate the ratios of the vacuum degree influence parameter and the mass axis offset parameter to the average vacuum degree influence parameter and the average mass axis offset parameter, and calculate the mean values to obtain the basic state anomaly coefficient;
[0082] Based on the verification error coefficient, the basic state anomaly coefficient is amplified and compensated to obtain the state anomaly coefficient, and fault identification and early warning are performed.
[0083] In this embodiment, a verification error coefficient is first calculated based on the vacuum verification coefficient and the mass verification coefficient, where the verification error coefficient = |vacuum verification coefficient - mass verification coefficient|. The vacuum verification coefficient represents the prediction accuracy of the vacuum influence analysis network; the higher the prediction accuracy, the larger the vacuum verification coefficient. The mass verification coefficient represents the prediction accuracy of the mass axis offset analysis network; the higher the prediction accuracy, the larger the mass verification coefficient. When the prediction accuracies of these two are similar, the verification error coefficient approaches 0, indicating that the model has strong predictive consistency in both the vacuum degree and mass axis offset dimensions. Furthermore, if the prediction accuracy of either the vacuum influence analysis network or the mass axis offset analysis network is abnormal, directly fusing the vacuum degree influence parameter and the mass axis offset parameter for fault warning may lead to distorted warning results due to single-dimensional prediction bias. For example, if the vacuum verification coefficient is 0.89 and the mass verification coefficient is 0.84, then the verification error coefficient = |0.89-0.84| = 0.05, indicating that the prediction accuracy of the vacuum influence analysis network and the mass axis offset analysis network is very synchronized at this time. If the vacuum verification coefficient is 0.89 and the mass verification coefficient is 0.44, then the verification error coefficient = |0.89-0.44| = 0.45, indicating that the prediction accuracy of the mass axis offset analysis network is low at this time. If the vacuum influence parameter and the mass axis offset parameter are directly integrated for fault warning, the warning result may be distorted due to the prediction deviation of the mass axis offset.
[0084] Secondly, the average vacuum level influence parameter and average mass axis offset parameter are obtained when the mass spectrometer experiences abnormal conditions. Specifically, based on the historical database, all vacuum level data and mass error data when the mass spectrometer experiences abnormal conditions are extracted. Then, all vacuum level influence parameters and mass axis offset parameters are calculated separately. Finally, the average vacuum level influence parameter and average mass axis offset parameter are obtained by averaging them. The average vacuum level influence parameter and average mass axis offset parameter reflect the vacuum level and mass axis offset levels under abnormal conditions in the historical data.
[0085] Next, calculate the ratios of the vacuum degree influence parameter and the mass axis offset parameter to the average vacuum degree influence parameter and the average mass axis offset parameter, and calculate the average values to obtain the basic state anomaly coefficient. The basic state anomaly coefficient is calculated as follows: (vacuum degree influence parameter / average vacuum degree influence parameter + mass axis offset parameter / average mass axis offset parameter) / 2. For example, at 25°C, the vacuum degree influence parameter is 0.2, the mass axis offset parameter is 0.25, the average vacuum degree influence parameter is 0.32, and the average mass axis offset parameter is 0.35. Therefore, the basic state anomaly coefficient is (0.2 / 0.32 + 0.25 / 0.35) / 2 = 0.67.
[0086] Finally, based on the verification error coefficient, the basic state anomaly coefficient is amplified and compensated to obtain the state anomaly coefficient, which is used for fault identification and early warning. The state anomaly coefficient is calculated as (1 + verification error coefficient) * basic state anomaly coefficient. Furthermore, the larger the difference between the vacuum verification coefficient and the mass verification coefficient, the more likely there are other hidden factors besides temperature causing the mass spectrometer's state anomaly. In this case, a larger verification error coefficient results in a greater amplification of the state anomaly coefficient, thus improving the fault identification and early warning response speed. For example, if the verification error coefficient is 0.05 and the basic state anomaly coefficient is 0.67, then the state anomaly coefficient is (1 + 0.05) * 0.67 = 0.7. This considers the consistency of prediction accuracy between the vacuum influence analysis network and the mass axis offset analysis network. When the prediction accuracy of either network is abnormal, the basic state anomaly coefficient is amplified and compensated using the verification error coefficient, thus avoiding distortion of the early warning result caused by single-dimensional prediction bias and enabling more accurate fault identification and early warning at an earlier stage.
[0087] In summary, compared to existing technologies, this application performs a fusion state anomaly analysis on vacuum degree influence parameters and mass axis offset parameters based on vacuum verification coefficients and mass verification coefficients to obtain state anomaly coefficients for fault identification and early warning. Thus, by compensating and amplifying the basic state anomaly coefficients through the absolute difference between the vacuum verification coefficients and mass verification coefficients, it avoids the distortion of early warning results caused by single-dimensional prediction bias, enabling more accurate fault identification and early warning at an earlier stage, facilitating maintenance and repair by relevant technical personnel.
[0088] In summary, the embodiments of this application have at least the following technical effects:
[0089] Compared with existing technologies, this application collects environmental parameters (including temperature) in the operating environment of the mass spectrometer, analyzes the impact of environmental parameters on the vacuum degree of the mass spectrometer, obtains vacuum degree impact parameters, and can evaluate the operating status of the mass spectrometer accordingly, providing a reliable basis for monitoring the operating status of the mass spectrometer.
[0090] Secondly, this application analyzes the impact of mass axis offset on the mass spectrometer based on environmental parameters to obtain mass axis offset parameters, which can be used to assess the operating status of the mass spectrometer and provide a reliable basis for subsequent results analysis.
[0091] Furthermore, this application obtains the actual vacuum level influence parameters and actual mass error parameters of the mass spectrometer through testing. These parameters, combined with the mass axis offset parameters, are used to verify the state influence, and the vacuum verification coefficient and mass verification coefficient are obtained through processing. In this way, the model prediction data is verified using actual data, and data support is provided for subsequent fault identification and early warning.
[0092] Finally, this application performs a fusion state anomaly analysis on the vacuum degree influence parameters and mass axis offset parameters based on the vacuum verification coefficient and mass verification coefficient to obtain a state anomaly coefficient for fault identification and early warning. In this way, the difference between the vacuum verification coefficient and the mass verification coefficient compensates for and amplifies the basic state anomaly coefficient, avoiding the distortion of early warning results caused by single-dimensional prediction bias. This allows for more accurate fault identification and early warning at an earlier stage, facilitating maintenance and repair by relevant technical personnel.
[0093] Through the above technical solution, this application uses a vacuum influence analysis network and a mass axis offset analysis network constructed based on machine learning. By taking environmental parameters as input, it outputs vacuum degree influence parameters and mass axis offset parameters, and compensates and amplifies the basic state anomaly coefficient accordingly. In this way, fault identification and early warning can be made at an earlier stage, improving the timeliness of mass spectrometer fault warning.
[0094] Example 2, as Figure 2As shown, based on the same inventive concept as the mass spectrometer operation status monitoring method provided in Embodiment 1, this embodiment of the invention also provides a mass spectrometer operation status monitoring system, including:
[0095] Vacuum analysis module 11 is used to collect environmental parameters in the operating environment of the mass spectrometer, and to perform vacuum degree influence analysis on the mass spectrometer based on the environmental parameters to obtain vacuum degree influence parameters.
[0096] The axis offset analysis module 12 is used to perform mass spectrometer mass axis offset influence analysis based on the environmental parameters and obtain mass axis offset parameters.
[0097] State verification module 13 is used to test and obtain the actual vacuum degree influence parameters and actual mass error parameters of the mass spectrometer, and to perform state influence verification by combining the vacuum degree influence parameters and mass axis offset parameters, and to process and obtain vacuum verification coefficient and mass verification coefficient.
[0098] The anomaly warning module 14 is used to perform a fusion state anomaly analysis on the vacuum degree influence parameter and the mass axis offset parameter based on the vacuum verification coefficient and the mass verification coefficient, obtain the state anomaly coefficient, and perform fault identification and warning.
[0099] Specifically, the vacuum analysis module 11 is used for:
[0100] Collect environmental parameters in the operating environment of the mass spectrometer, including ambient temperature;
[0101] The environmental parameters are input into the vacuum influence analysis network of the mass spectrometer, and the vacuum influence parameters are identified and output.
[0102] Furthermore, the training steps for the "vacuum influence analysis network" include:
[0103] Based on the operation and monitoring data of the mass spectrometer, a set of sample environmental parameters was collected, and the vacuum degree change parameters of the mass spectrometer under different sample environmental parameters were collected. The set of parameters affecting the sample vacuum degree was then labeled.
[0104] Construct a machine learning-based network for vacuum effect analysis;
[0105] Under the set of sample environmental parameters and the set of sample vacuum degree influence parameters, the vacuum influence analysis network is trained and the parameters are optimized under supervision. The training is completed after the accuracy is tested and converged.
[0106] Specifically, the axis deviation analysis module 12 is used for:
[0107] The mass axis offset analysis network is invoked, wherein the mass axis offset analysis network is trained using a set of sample environment parameters and a set of sample mass axis offset parameters, and the sample mass axis offset parameters include the mass error magnitude;
[0108] The environmental parameters are input into the mass axis offset analysis network, and the mass axis offset parameters are obtained from the output.
[0109] Specifically, the status verification module 13 is used for:
[0110] The actual vacuum level and actual mass error parameters of the mass spectrometer were obtained through testing.
[0111] Based on the vacuum degree influence parameters and the actual vacuum degree influence parameters, the vacuum state influence is verified to obtain the vacuum verification coefficient.
[0112] The impact of the quality status is verified based on the actual quality error parameters and the quality axis offset parameters to obtain the quality verification coefficient.
[0113] Furthermore, the step of "verifying the vacuum state influence based on the vacuum degree influence parameters and the actual vacuum degree influence parameters, and obtaining the vacuum verification coefficient" includes:
[0114] Calculate the vacuum deviation amplitude between the vacuum degree influence parameters and the actual vacuum degree influence parameters;
[0115] The vacuum verification coefficient is calculated based on the vacuum deviation amplitude.
[0116] The anomaly warning module 14 is specifically used for:
[0117] The verification error coefficient is calculated based on the vacuum verification coefficient and the mass verification coefficient.
[0118] Obtain the average vacuum level and average mass axis offset parameters when the mass spectrometer experiences abnormal conditions.
[0119] Calculate the ratios of the vacuum degree influence parameter and the mass axis offset parameter to the average vacuum degree influence parameter and the average mass axis offset parameter, and calculate the mean values to obtain the basic state anomaly coefficient;
[0120] Based on the verification error coefficient, the basic state anomaly coefficient is amplified and compensated to obtain the state anomaly coefficient, and fault identification and early warning are performed.
[0121] In summary, the embodiments of this application have at least the following technical effects:
[0122] Compared to existing technologies, this application first uses a vacuum analysis module to collect environmental parameters in the mass spectrometer's operating environment. Based on these parameters, it analyzes the impact of vacuum on the mass spectrometer, obtaining vacuum influence parameters. This allows for the assessment of the mass spectrometer's operating status, providing a reliable basis for monitoring its operational status. Secondly, using an axis offset analysis module, it analyzes the impact of mass axis offset on the mass spectrometer based on environmental parameters, obtaining mass axis offset parameters. This provides a reliable basis for subsequent results analysis. Thirdly, using a state verification module, it tests and obtains the actual vacuum influence parameters and actual mass error parameters of the mass spectrometer. Combining these parameters with the mass axis offset parameters, it verifies the state impact, processing them to obtain vacuum verification coefficients and mass verification coefficients. This verifies the model's predicted data using actual data and provides data support for subsequent fault identification and early warning. Finally, the anomaly warning module performs a fusion state anomaly analysis on vacuum degree influence parameters and mass axis offset parameters based on vacuum verification coefficients and mass verification coefficients to obtain state anomaly coefficients. This allows for fault identification and early warning. The absolute difference between the vacuum verification coefficient and the mass verification coefficient is used to compensate for and amplify the basic state anomaly coefficient, avoiding distortion of warning results caused by single-dimensional prediction bias. This enables more accurate fault identification and early warning at an earlier stage, facilitating maintenance and repair by relevant technical personnel. This improves the timeliness of mass spectrometer fault warnings.
[0123] Example 3, as Figure 3 As shown, an embodiment of the present invention provides an electronic device 200, including a memory 210, a processor 220, and a computer program 211 stored in the memory 210 and executable on the processor 220. When the processor 220 executes the computer program 211, it implements a mass spectrometer operation status monitoring method in Embodiment 1.
[0124] Example 4: This example provides a computer-readable storage medium storing a computer program. When the computer program is executed by a processor, it implements a mass spectrometer operation status monitoring method as described in Example 1.
[0125] It should be noted that the descriptions of each embodiment in the above embodiments have different focuses. For parts that are not described in detail in a certain embodiment, please refer to the relevant descriptions in other embodiments.
[0126] Those skilled in the art will understand that embodiments of the present invention can be provided as methods, systems, or computer program products. Therefore, the present invention can take the form of a completely hardware embodiment, a completely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, the present invention can take the form of a computer program product embodied on one or more computer-usable storage media (including, but not limited to, disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.
[0127] This invention is described with reference to flowchart illustrations and / or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of the invention. It will be understood that each block of the flowchart illustrations and / or block diagrams, and combinations of blocks in the flowchart illustrations and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, special-purpose computer, embedded computer, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, generate instructions for implementing the flowchart illustrations. Figure 1 One or more processes and / or boxes Figure 1 A device that provides the functions specified in one or more boxes.
[0128] These computer program instructions may also be stored in a computer-readable storage medium that can direct a computer or other programmable data processing device to function in a particular manner, such that the instructions stored in the computer-readable storage medium produce an article of manufacture including instruction means, which are implemented in a process Figure 1 One or more processes and / or boxes Figure 1 The function specified in one or more boxes.
[0129] These computer program instructions may also be loaded onto a computer or other programmable data processing equipment to cause a series of operational steps to be performed on the computer or other programmable equipment to produce a computer-implemented process, thereby providing instructions that execute on the computer or other programmable equipment for implementing the process. Figure 1 One or more processes and / or boxes Figure 1 The steps of the function specified in one or more boxes.
[0130] Although preferred embodiments of the invention have been described, those skilled in the art, once they have learned the basic inventive concept, can make other changes and modifications to these embodiments.
[0131] Obviously, those skilled in the art can make various modifications and variations to this invention without departing from its spirit and scope. Therefore, if these modifications and variations fall within the scope of this invention and its equivalents, this invention also intends to include these modifications and variations.
Claims
1. A method for monitoring the operating status of a mass spectrometer, characterized in that, The method includes: Environmental parameters in the operating environment of the mass spectrometer are collected, and the influence of the vacuum degree of the mass spectrometer is analyzed based on the environmental parameters to obtain the vacuum degree influence parameters. Based on the environmental parameters, an analysis of the impact of mass spectrometer mass axis offset was performed to obtain mass axis offset parameters. The actual vacuum degree influence parameters and actual mass error parameters of the mass spectrometer are obtained through testing. The state influence is verified by combining the vacuum degree influence parameters and the mass axis offset parameters. The vacuum verification coefficient and mass verification coefficient are then obtained through processing. Based on the vacuum verification coefficient and the mass verification coefficient, a fusion state anomaly analysis is performed on the vacuum degree influence parameter and the mass axis offset parameter to obtain the state anomaly coefficient and perform fault identification and early warning. This involves collecting environmental parameters from the mass spectrometer's operating environment, analyzing the impact of these environmental parameters on the mass spectrometer's vacuum level, and obtaining vacuum level impact parameters, including: Collect environmental parameters in the operating environment of the mass spectrometer, including ambient temperature; The environmental parameters are input into the vacuum effect analysis network of the mass spectrometer, and the vacuum effect parameters are identified and output. The analysis of the impact of the mass spectrometer mass axis shift based on the environmental parameters yields the mass axis shift parameters, including: The mass axis offset analysis network is invoked, wherein the mass axis offset analysis network is trained using a set of sample environment parameters and a set of sample mass axis offset parameters, and the sample mass axis offset parameters include the mass error magnitude; The environmental parameters are input into the mass axis offset analysis network, and the mass axis offset parameters are obtained by identifying the output. The process involves testing and obtaining the actual vacuum level influence parameters and actual mass error parameters of the mass spectrometer. These parameters, combined with the vacuum level influence parameters and mass axis offset parameters, are used to verify the state influence. The resulting data are then processed to obtain vacuum verification coefficients and mass verification coefficients, including: The actual vacuum level and actual mass error parameters of the mass spectrometer were obtained through testing. Based on the vacuum degree influence parameters and the actual vacuum degree influence parameters, the vacuum state influence is verified to obtain the vacuum verification coefficient. The impact of the quality status is verified based on the actual quality error parameters and the quality axis offset parameters to obtain the quality verification coefficient.
2. The method for monitoring the operating status of a mass spectrometer according to claim 1, characterized in that, The training steps for the vacuum influence analysis network include: Based on the operation and monitoring data of the mass spectrometer, a set of sample environmental parameters was collected, and the vacuum degree change parameters of the mass spectrometer under different sample environmental parameters were collected. The set of parameters affecting the sample vacuum degree was then labeled. Construct a machine learning-based network for vacuum effect analysis; Under the set of sample environmental parameters and the set of sample vacuum degree influence parameters, the vacuum influence analysis network is trained and the parameters are optimized under supervision. The training is completed after the accuracy is tested and converged.
3. The method for monitoring the operating status of a mass spectrometer according to claim 1, characterized in that, Based on the vacuum degree influence parameters and the actual vacuum degree influence parameters, the vacuum state influence is verified to obtain the vacuum verification coefficient, including: Calculate the vacuum deviation amplitude between the vacuum degree influence parameters and the actual vacuum degree influence parameters; The vacuum verification coefficient is calculated based on the vacuum deviation amplitude.
4. The method for monitoring the operating status of a mass spectrometer according to claim 1, characterized in that, Based on the vacuum verification coefficient and the mass verification coefficient, a fusion state anomaly analysis is performed on the vacuum degree influence parameter and the mass axis offset parameter to obtain a state anomaly coefficient, and fault identification and early warning are performed, including: The verification error coefficient is calculated based on the vacuum verification coefficient and the mass verification coefficient. Obtain the average vacuum level and average mass axis offset parameters when the mass spectrometer experiences abnormal conditions. Calculate the ratios of the vacuum degree influence parameter and the mass axis offset parameter to the average vacuum degree influence parameter and the average mass axis offset parameter, and calculate the mean values to obtain the basic state anomaly coefficient; Based on the verification error coefficient, the basic state anomaly coefficient is amplified and compensated to obtain the state anomaly coefficient, and fault identification and early warning are performed.
5. A mass spectrometer operating status monitoring system, characterized in that, For performing the method according to any one of claims 1-4, comprising: The vacuum analysis module is used to collect environmental parameters in the operating environment of the mass spectrometer, and to perform vacuum degree influence analysis based on the environmental parameters to obtain vacuum degree influence parameters. The axis offset analysis module is used to analyze the impact of the mass spectrometer mass axis offset based on the environmental parameters, and to obtain the mass axis offset parameters. The state verification module is used to test and obtain the actual vacuum degree influence parameters and actual mass error parameters of the mass spectrometer, and to perform state influence verification by combining the vacuum degree influence parameters and mass axis offset parameters, and to process and obtain vacuum verification coefficient and mass verification coefficient. The anomaly warning module is used to perform a fusion state anomaly analysis on the vacuum degree influence parameter and the mass axis offset parameter based on the vacuum verification coefficient and the mass verification coefficient, obtain the state anomaly coefficient, and perform fault identification and warning.
6. An electronic device, characterized in that, include: Memory, used to store computer software programs; A processor is used to read and execute the computer software program, thereby implementing the mass spectrometer operation status monitoring method according to any one of claims 1-4.
7. A computer-readable storage medium, characterized in that, The storage medium stores a computer program, which, when executed by a processor, implements a mass spectrometer operation status monitoring method as described in any one of claims 1-4.