Neutron Chopper Operation and Maintenance Health Status Monitoring System

Abstract

This invention relates to a neutron chopper operation and maintenance health status monitoring system, comprising a data acquisition module installed on the chopper and a database management module, a status display module, and a fault diagnosis module at a remote terminal. The data acquisition module is used to collect vibration, current, temperature, and vacuum data of the chopper. The database management module is used to receive the data collected by the data acquisition module and store the received data in the database. The fault diagnosis module is used to diagnose whether the data stored in the database is faulty data and feeds back the diagnosis results to the database management module. The status display module reads the latest monitoring data from the database and displays the data and fault information to the user.

Description

Technical Field

[0001] This invention relates to the field of neutron chopper monitoring, and more particularly to a neutron chopper operation and maintenance health status monitoring system. Background Technology

[0002] A neutron chopper (large scientific facility chopper) is a mechanical chopper belonging to the category of rotating machinery. Also known as a neutron energy selector, it emerged in the 1930s and 40s and was initially used for the study of neutron-induced radioactive isotopes. With the development of neutron scattering technology, it has been widely used in various large-scale neutron source facilities. In time-of-flight neutron scattering spectrometers, the chopper can time-cut the beam, thereby removing fast neutrons and high-energy gamma-ray backgrounds, selecting neutron wavelength ranges, and monochromating neutrons.

[0003] Neutron choppers exhibit good stability during actual operation and maintenance, so health status monitoring is generally not required. However, neutron choppers can still experience certain faults during actual operation and maintenance. The main fault factors include vibration, temperature, vacuum level, and current. Therefore, it is urgent to design a neutron chopper operation and maintenance health status monitoring system to monitor and provide feedback on the operating status of the neutron chopper in real time. Summary of the Invention

[0004] The purpose of this invention is to provide a neutron chopper operation and maintenance health status monitoring system. The system collects the chopper's operating status periodically through a data acquisition module, writes the collected data into a database through a database management module, determines whether the chopper has malfunctioned based on the data in the database, and displays the latest monitoring data and fault information to the user through a status display module.

[0005] To achieve the above objectives, the technical solution adopted in this invention is as follows: a neutron chopper operation and maintenance health status monitoring system, including a data acquisition module installed on the chopper and a database management module, a status display module, and a fault diagnosis module at a remote terminal. The data acquisition module is used to collect vibration, current, temperature, and vacuum data of the chopper; the database management module is used to receive the data collected by the data acquisition module and store it in the database; the fault diagnosis module is used to diagnose whether the data recorded in the database is fault data and to feed back the diagnosis results to the database management module; the status display module is used to display the collected data information and fault information.

[0006] Preferably, the diagnostic steps of the fault diagnosis module are as follows:

[0007] Step 1: Extract effective features from normal state data using the generative adversarial mechanism of the Bi-GAN algorithm;

[0008] Step 2: Construct a fault detection model based on a single-class learning algorithm using the effective features extracted in Step 1;

[0009] Step 3: After the data from the database is processed by the feature extractor, it is input into the fault detection model to determine whether it is fault data, and the diagnosis results are fed back to the database management module.

[0010] Preferably, the unit learning algorithm used in step two is Support Vector Machine (OC-SVM), and the basic function of Support Vector Machine is given by the following formula:

[0011] f(x) = ω·∮(x) + b

[0012] Where f(x) is a nonlinear mapping function, and ω and b are vector parameters that need to be determined; in the dataset {(x1, y1), (x2, y2), ..., (x... n y n )}(x i ∈R n y i ∈R n ) in x i For input, y i It is the output, and R n Representing an n-dimensional Euclidean space; a trainable function f(x) approximates the actual value y; the objective function of the support vector machine can be expressed as:

[0013]

[0014] Where ∫ is the cost function, C is the penalty factor, and n represents the number of data points;

[0015] The cost function of SVM is as follows:

[0016]

[0017] Where ε is a parameter; the SVM method places the maximum deviation ε between the predicted value f(x) and the actual value y between the two, and starts calculating the cost value when the absolute difference between the two is greater than ε; the two-width margin 2ε will be built below the center of f(x); the right boundary point is the deterministic model of the “support vector”; if the training samples remain in the blank, the result can be assumed to be correct.

[0018] Due to the flexible boundary, the slack variable ξ i and Therefore, the equation can be rewritten as:

[0019]

[0020]

[0021] In the formula, ξ i and Let represent the upper and lower bounds of the error, respectively; by introducing Lagrange multipliers α≥0, δ≥0; the Lagrange function is:

[0022]

[0023] The dual function can be written as:

[0024]

[0025]

[0026] function K(x) i ,y j () is the kernel function. The kernel function can transform a nonlinear problem in low-dimensional space into a linear programming problem in high-dimensional space. The approximate function is:

[0027]

[0028] Preferably, step one utilizes a bidirectional generative adversarial network (Bi-GAN) algorithm, which is an improvement on the GAN algorithm. It not only includes a generator G (which only contains a generator from the feature space to the data space) but also a feature extractor E (encoder) (which extracts features from the data space to the feature space). The discriminator determines whether the data pair (x, z) follows a joint distribution p. EX (x, z) = p X (x)p E (z|x) or: p GZ (x, z) = p G (x|z)p Z (z).

[0029] Preferably, the data acquisition module includes a current sensor for acquiring chopper current signals, a temperature sensor for acquiring chopper internal temperature signals, a vacuum sensor for acquiring chopper internal vacuum signals, and a vibration sensor for acquiring chopper vibration signals; an external block is provided on the outer side of the chopper, and two external holes are provided on the external block. The two external holes are fitted with the temperature sensor and the vacuum sensor that can be inserted into the chopper, and the temperature sensor and the vacuum sensor achieve sealing of the external holes after installation.

[0030] Preferably, the chopper is equipped with four mounting blocks. A right-angled vibration detection seat is set on the mounting block. The vertical side of the vibration detection seat is attached to the housing of the chopper, and a vibration sensor is installed on the vertical side through a mounting flange. The horizontal side is set on the mounting block, and the vibration detection seat has a stepped adjustment hole with a larger upper part and a smaller lower part. The lower part of the adjustment hole matches the mounting hole on the mounting block, and the lower part of the adjustment hole is larger than the mounting hole on the mounting block. A locking nut is placed on the upper part of the adjustment hole.

[0031] The technical effects of this invention are as follows:

[0032] 1. The data acquisition module periodically collects the operating status of the chopper and transmits the collected data to the database management module. The database management module writes the data into the database. The fault diagnosis module is used to determine whether the data in the database is faulty data. The status display module displays the latest monitoring data and fault information to the user. In this way, the operation and maintenance status monitoring of the neutron chopper can be realized.

[0033] 2. The design of the external block allows the temperature sensor and vacuum sensor to be mounted outside the chopper, making them easy to install and remove. At the same time, the sealed installation will not affect the vacuum level of the chopper.

[0034] 3. The design of the adjustment hole allows for fine-tuning of the vibration detection seat's position, ensuring that the vibration detection seat is tightly fitted to the chopper's housing. This ensures that the vibration of the chopper's housing directly drives the vertical side of the vibration detection seat to vibrate, thereby enabling the vibration sensor to accurately measure the vibration amplitude. Furthermore, the chopper and vibration sensor are directly mounted using a single bolt, ensuring that their mounting positions are consistent and guaranteeing that the vibration signal collected by the vibration detector is the chopper's own vibration, unaffected by the vibration of the mounting bracket or seat. Attached Figure Description

[0035] Figure 1 This is a schematic diagram of a neutron chopper operation and maintenance health status monitoring system.

[0036] Figure 2 This is a schematic diagram of the data acquisition module.

[0037] Figure 3 This is a schematic diagram illustrating the single-class data feature extraction using the Bi-GAN (Bi-Generative Adversarial Network) algorithm.

[0038] Figure 4 This is a three-dimensional schematic diagram of a chopper.

[0039] Figure 5 This is a schematic diagram showing the interaction between the vibration sensor and the chopper.

[0040] Figure 6This is a schematic diagram of the chopper after being fitted with the locking bolt.

[0041] The text labels in the diagram represent: 1. Chopper; 2. Mounting block; 3. External block; 4. External hole; 5. Vibration detection seat; 6. Vibration sensor; 7. Mounting flange; 8. Adjustment hole; 9. Locking nut; 10. Locking bolt. Detailed Implementation

[0042] To enable those skilled in the art to better understand the technical solution of the present invention, the present invention will be described in detail below. The description in this part is only exemplary and explanatory, and should not be used to limit the scope of protection of the present invention in any way.

[0043] like Figure 1-3 As shown, the technical solution of this application is: a neutron chopper operation and maintenance health status monitoring system, characterized in that it includes a data acquisition module installed on the chopper and a database management module, a status display module, and a fault diagnosis module at a remote terminal. The data acquisition module is used to collect vibration, current, temperature, and vacuum data of the chopper 1; the database management module is used to receive the data collected by the data acquisition module and store the data in the database; the fault diagnosis module is used to diagnose whether the data recorded in the database is fault data; the status display module is used to display the collected data information and fault information. The diagnostic steps of the fault diagnosis module are as follows:

[0044] Step 1: Extract effective features from normal state data using the generative adversarial mechanism of the Bi-GAN algorithm;

[0045] Step 2: Construct a fault detection model based on a single-class learning algorithm using the effective features extracted in Step 1;

[0046] Step 3: After the data from the database is processed by the feature extractor, it is input into the fault detection model to determine whether it is fault data, and the diagnosis results are fed back to the database management module.

[0047] The single-class learning algorithm used in step two is Support Vector Machine (OC-SVM), and the basic function of Support Vector Machine is given by the following formula:

[0048] f(x)=ω·∮(x)+ b

[0049] Where f(x) is a nonlinear mapping function, and ω and b are vector parameters that need to be determined; in the dataset {(x1, y1), (x2, y2), ..., (x... n y n )}(x i ∈R n y i ∈R n) in x i For input, y i It is the output, and R n Representing an n-dimensional Euclidean space; a trainable function f(x) approximates the actual value y; the objective function of the support vector machine can be expressed as:

[0050]

[0051] Where ∫ is the cost function, C is the penalty factor, and n represents the number of data points;

[0052] The cost function of SVM is as follows:

[0053]

[0054] Where ε is a parameter; the SVM method places the maximum deviation ε between the predicted value f(x) and the actual value y between the two, and starts calculating the cost value when the absolute difference between the two is greater than ε; the two-width margin 2ε will be built below the center of f(x); the right boundary point is the deterministic model of the “support vector”; if the training samples remain in the blank, the result can be assumed to be correct.

[0055] Due to the flexible boundary, the slack variable ξ i and Therefore, the equation can be rewritten as:

[0056]

[0057]

[0058] In the formula, ξ i and Let represent the upper and lower bounds of the error, respectively; by introducing Lagrange multipliers α≥0, δ≥0; the Lagrange function is:

[0059]

[0060] The dual function can be written as:

[0061]

[0062]

[0063] function K(x) i ,y j () is the kernel function. The kernel function can transform a nonlinear problem in low-dimensional space into a linear programming problem in high-dimensional space. The approximate function is:

[0064]

[0065] Step one utilizes the Bidirectional Generative Adversarial Network (Bi-GAN) algorithm, an improvement on the GAN algorithm. It not only incorporates a generator G (which only contains a generator to transform the data space from the feature space) but also includes a feature extractor E (Encoder) to transform the data space into the feature space. The discriminator determines whether the data pair (x, z) follows a joint distribution p. EX (x, z) = p X (x)p E (z|x) or: p GZ (x, z) = p G (x|z)p Z (z).

[0066] The specific application is as follows: First, data under normal operating conditions is written to the database management module. Temperature and vacuum levels are within specific ranges; within these ranges, the data is considered normal, and outside, it is considered abnormal. These two indicators do not require a fault detection model. However, vibration and current require separate fault detection models. Specifically, effective features are extracted from the normal state data using the generative adversarial mechanism of the Bidirectional Generative Adversarial Network (Bi-GAN) algorithm. The extracted effective features are then used to build a fault detection model based on the Support Vector Machine (OC-SVM) algorithm. Next, the data acquisition module periodically collects temperature, vacuum, current, and vibration data of the chopper and transmits the collected data to the database management system for storage. The fault diagnosis module reads the data from the database, processes the data through a feature extractor, and inputs it into the fault detection model to determine whether it is faulty data. The diagnosis result is then fed back to the database management module. The status display module displays the latest monitoring data and fault information, thus realizing the status monitoring of the neutron chopper's operation and maintenance. To better verify the fault detection model, manual verification can be performed after a fault is detected, and the verification result is then updated in the database management module.

[0067] like Figure 4-6As shown, the data acquisition module includes a current sensor for acquiring chopper current signals, a temperature sensor for acquiring chopper internal temperature signals, a vacuum sensor for acquiring chopper internal vacuum signals, and a vibration sensor for acquiring chopper vibration signals. An external block 3 is provided on the outer side of the chopper 1. Two external holes 4 are provided on the external block 3, and the temperature sensor and vacuum sensor, which can be inserted into the chopper 1, are fitted into the two external holes 4. The installed temperature sensor and vacuum sensor seal the external holes 4. The chopper 1 is equipped with four mounting blocks 2. A right-angled vibration detection seat 5 is provided on the mounting block 2. The vertical edge of the vibration detection seat 5 is attached to the housing of the chopper 1, and a vibration sensor 6 is installed on the vertical edge via a mounting flange 7. The horizontal edge is located on the mounting block 2, and the vibration detection seat 5 has a stepped adjustment hole 8 with a larger upper part and a smaller lower part. The lower part of the adjustment hole 8 mates with a mounting hole on the mounting block 2, and the lower part of the adjustment hole 8 is larger than the mounting hole on the mounting block 2. A locking nut 9 is placed on the upper part of the adjustment hole 8.

[0068] The current sensor is typically installed inside the neutron chopper. During installation, other components are secured by a locking bolt 10 passing through the mounting hole and the upper part of the adjustment hole 8 on the mounting block 2, and then through the locking nut 9. The mounting block 2 is then secured to the mounting base or bracket by the locking bolt 10 and the locking nut 9. The adjustment hole allows for fine-tuning of the position of the vibration detection seat 5, ensuring it fits snugly against the chopper 1's housing. This ensures that the vibration of the chopper 1's housing directly drives the vertical side of the vibration detection seat 5 to vibrate, thereby enabling the vibration sensor 6 to... The vibration amplitude is accurately measured, and the chopper and data acquisition unit are installed directly with a single bolt. This ensures that the chopper and data acquisition unit are locked in the same position, guaranteeing that the vibration signal collected by the vibration sensor 6 is the vibration of the chopper 1 itself and is not affected by the vibration of the mounting bracket or mounting base. The external hole 4 is generally a threaded hole and communicates with the inside of the chopper. The external block 3 is generally located on the opposite side of the chopper's cooling fan. The temperature sensor and vacuum sensor are installed in the external hole 4 by threads and sealed by a flange.

[0069] It should be noted that, in this document, the terms “comprising,” “including,” or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such process, method, article, or apparatus.

[0070] This article uses specific examples to illustrate the principles and implementation methods of the present invention. The above examples are only for the purpose of helping to understand the method and core ideas of the present invention. The above descriptions are only preferred embodiments of the present invention. It should be noted that due to the limitations of textual expression, while there are objectively infinite specific structures, those skilled in the art can make several improvements, modifications, or changes without departing from the principles of the present invention, and can also combine the above technical features in an appropriate manner. These improvements, modifications, changes, or combinations, or the direct application of the inventive concept and technical solution to other situations without modification, should all be considered within the scope of protection of the present invention.

Claims

1. A neutron chopper operation and maintenance health status monitoring system, characterized in that, It includes a data acquisition module installed on chopper 1 and a database management module, status display module, and fault diagnosis module at a remote terminal. The data acquisition module is used to collect vibration, current, temperature, and vacuum data of chopper 1. The database management module is used to receive the data collected by the data acquisition module and store the data in the database. The fault diagnosis module is used to diagnose whether the data recorded in the database is faulty data and to feed back the diagnosis results to the database management module. The status display module is used to display the collected data information and fault information. The diagnostic steps of the fault diagnosis module are as follows: Step 1: Extract effective features from normal state data using the generative adversarial mechanism of the bidirectional generative adversarial network algorithm; Step 2: Construct a fault detection model based on a single-class learning algorithm using the effective features extracted in Step 1; Step 3: After the data from the database is processed by the feature extractor, it is input into the fault detection model to determine whether it is fault data, and the diagnosis results are fed back to the database management module.

2. The neutron chopper operation and maintenance health status monitoring system according to claim 1, characterized in that, The single-class learning algorithm used in step two is the Support Vector Machine (SVM). The basic function of the SVM is given by the following equation: f(x) = ω·∮(x) + b Where f(x) is a nonlinear mapping function, and ω and b are vector parameters that need to be determined; in the dataset {(x 1， y1), (x 2， y2), ...,(x n， y n )}(x i ∈R n y i ∈R n ) in x i For input, y i It is the output, and R n Representing an n-dimensional Euclidean space; a trainable function f(x) approximates the actual value y; the objective function of the support vector machine can be expressed as: Where ∫ is the cost function, C is the penalty factor, and n represents the number of data points; The cost function of SVM is as follows: Where ε is a parameter; the SVM method places the maximum deviation ε between the predicted value f(x) and the actual value y between the two, and starts calculating the cost value when the absolute difference between the two is greater than ε; the two-width margin 2ε will be built below the center of f(x); the right boundary point is the deterministic model of the "support vector"; if the training samples remain in the blank, the result can be assumed to be correct; Due to the flexible boundary, the slack variable ξ i and Therefore, the equation can be rewritten as: In the formula, ξ i and Let represent the upper and lower bounds of the error, respectively; by introducing Lagrange multipliers α≥0, δ≥0; the Lagrange function is: The dual function can be written as: function K(x) i ,y j () is the kernel function. The kernel function can transform a nonlinear problem in low-dimensional space into a linear programming problem in high-dimensional space. The approximate function is:

3. The neutron chopper operation and maintenance health status monitoring system according to claim 1, characterized in that, Step one utilizes a bidirectional generative adversarial network (GAN) algorithm, which is an improvement on the GAN algorithm. It not only incorporates a generator G (which only contains a generator from the feature space to the data space) but also a feature extractor E (which extracts features from the data space to the feature space). The discriminator determines whether the data pair (x,z) follows a joint distribution p. EX (x, z) = p x (x)p E (z|x) or p GZ (x, z) = p G (x|z)p Z (z).

4. The neutron chopper operation and maintenance health status monitoring system according to claim 1, characterized in that, The data acquisition module includes a current sensor for acquiring chopper current signals, a temperature sensor for acquiring chopper internal temperature signals, a vacuum sensor for acquiring chopper internal vacuum signals, and a vibration sensor for acquiring chopper vibration signals. An external block 3 is provided on the outer side of the chopper 1. Two external holes 4 are provided on the external block 3. The temperature sensor and vacuum sensor can be inserted into the chopper 1 through the two external holes 4, and the temperature sensor and vacuum sensor are sealed after installation.

5. The neutron chopper operation and maintenance health status monitoring system according to claim 4, characterized in that, The chopper 1 is equipped with four mounting blocks 2. A right-angled vibration detection seat 5 is set on the mounting block 2. The vertical side of the vibration detection seat 5 is attached to the housing of the chopper 1, and a vibration sensor 6 is installed on the vertical side through the mounting flange 7. The horizontal side is set on the mounting block 2, and the vibration detection seat 5 has a stepped adjustment hole 8 with a larger upper part and a smaller lower part. The lower part of the adjustment hole 8 matches the mounting hole on the mounting block 2, and the lower part of the adjustment hole 8 is larger than the mounting hole on the mounting block 2. A locking nut 9 is placed on the upper part of the adjustment hole 8.

Images