Centrifugal releaser safety performance detection system

Abstract

The invention relates to the technical field of centrifugal releaser detection, and discloses a centrifugal releaser safety performance detection system, which comprises a data acquisition module, a data preprocessing module, a data interaction module, a data fusion module, a diagnosis module and an output module, through deep application of multi-module collaborative design and frontier technology, the detection system realizes comprehensive improvement of detection performance, the data acquisition module utilizes a multi-sensor array and a time synchronization algorithm to ensure acquisition of multi-dimensional data of operation of the centrifugal releaser, and a foundation is built for accurate detection; the preprocessing module efficiently removes data noise, extracts key features and relieves follow-up processing pressure by means of Kalman filtering and short-time Fourier transform, the data interaction module achieves rapid data transmission and local processing, reduces network loads and guarantees real-time performance based on the Internet of Things and edge calculation, and the designs are linked with one another, so that the data interaction efficiency is improved. Therefore, the system can quickly and accurately acquire and process data, and the detection efficiency is greatly improved.

Description

Technical Field

[0001] The present invention relates to the technical field of centrifugal releaser detection, and more particularly discloses a centrifugal releaser safety performance detection system. Background Art

[0002] The centrifugal release is a device that works on the principle of centrifugal force and is primarily used in braking systems. With its simple structure, compact size, flexibility, and reliability, the centrifugal release is widely used in the safety braking systems of auxiliary transportation equipment in mines. Safety performance testing of centrifugal releases is necessary for several reasons: ensuring the normal operation and safety of the equipment, preventive maintenance, and ensuring production continuity and stability. Existing detection systems use only a single sensor to collect data, making it impossible to obtain comprehensive equipment operating data. This can easily lead to deviations in detection results due to missing data. Furthermore, the collected data is directly transmitted without preprocessing, which not only increases network pressure but also reduces processing efficiency. Simulating operating conditions relies solely on simple physical tests or virtual simulations, making it difficult to restore complex scenarios and disconnecting detection results from actual operating conditions. Finally, the lack of a dynamic weight allocation mechanism makes it difficult to eliminate differences in multi-source data, or relies on manual experience and simple algorithms, making it impossible to automatically and accurately identify potential faults. Summary of the Invention

[0003] The main technical problem solved by the present invention is to provide a centrifugal releaser safety performance detection system, which can solve the problems existing in the above-mentioned background technology.

[0004] To solve the above technical problems, according to one aspect of the present invention, more specifically, a centrifugal releaser safety performance detection system comprises: S1: using a data acquisition module to comprehensively collect operating data of the centrifugal releaser using multiple types of sensors to ensure that various types of information when the equipment is working can be obtained, providing a rich and comprehensive data foundation for subsequent detection, and avoiding inaccurate detection results due to missing data; S2: The preprocessing module performs preliminary processing on the collected data to remove noise and redundant information. This not only reduces the pressure on data transmission but also quickly extracts key data features, significantly improving data processing efficiency and reducing the computational burden of subsequent modules. S3: The data interaction module acts as a bridge, enabling smooth data flow between different modules. This allows data to be transmitted to where it is needed in a timely and accurate manner, ensuring the coordinated operation of all parts of the system and enhancing the integrity and coherence of the system. S4: The simulation module simulates the operating status of the centrifugal releaser under various actual working conditions and verifies it with real data. This makes up for the shortcomings of relying solely on actual testing or virtual simulation, greatly improving the authenticity and accuracy of working condition simulation, so that the test results are more in line with the actual operation of the equipment; S5: The data fusion module deeply integrates the pre-processed virtual simulation data with the physical simulation data, eliminating differences and contradictions between the data, unifying the data format and standards, and providing more accurate and reliable data support for subsequent analysis; S6: The diagnostic module uses intelligent algorithms to analyze fused data, automatically identify potential equipment failures and generate diagnostic reports.

[0005] Furthermore, the simulation module includes: a virtual simulation module and a physical simulation module; Virtual simulation module: Use computer simulation software ANSYS to build a three-dimensional virtual model of the centrifugal releaser, input actual physical parameters and working conditions, and perform simulation analysis; Physical simulation module: Using variable frequency motors, programmable controllers and hydraulic systems, we build a physical test platform that can simulate different slopes, speeds and loads to obtain operating data in real environments.

[0006] Furthermore, the data acquisition module adopts a multi-sensor array, including a MEMS acceleration sensor, a Hall speed sensor and a piezoresistive pressure sensor, and each sensor realizes precise triggering of data acquisition through a time synchronization algorithm.

[0007] Furthermore, the pre-processing module integrates Kalman filter algorithm and short-time Fourier transform to perform denoising and time-frequency domain feature extraction on the original data to achieve data dimensionality reduction.

[0008] Furthermore, the data fusion module adopts a dynamic weight allocation mechanism to adjust the fusion weight of sensor data and virtual simulation data in real time according to working condition parameters, and realizes synchronous processing of multi-source data through a timestamp alignment algorithm.

[0009] Furthermore, the diagnosis module constructs a fault recognition model based on convolutional neural networks and long short-term memory networks.

[0010] Furthermore, the result output module presents the test results in the form of visual charts and text descriptions based on the diagnosis report generated by the diagnosis module.

[0011] The beneficial effects of the centrifugal releaser safety performance detection system of the present invention are as follows: through multi-module collaborative design and in-depth application of cutting-edge technologies, the detection system achieves a comprehensive improvement in detection performance; the data acquisition module uses a multi-sensor array and a time synchronization algorithm to ensure the acquisition of multi-dimensional data on the operation of the centrifugal releaser, laying a solid foundation for accurate detection; the preprocessing module uses Kalman filtering and short-time Fourier transform to efficiently remove data noise, extract key features, and reduce subsequent processing pressure; the data interaction module is based on the Internet of Things and edge computing to achieve rapid data transmission and local processing, reduce network load and ensure real-time performance. These designs are linked together, enabling the system to quickly and accurately acquire and process data, greatly improving detection efficiency.

[0012] Through the organic combination of virtual simulation and physical simulation, as well as the innovative application of intelligent algorithms, the system's detection accuracy and reliability have been significantly enhanced. In the simulation module, virtual simulation and physical simulation verify each other, highly restoring the operating status of the equipment under complex working conditions; the data fusion module uses dynamic weight distribution and timestamp alignment to eliminate differences in multi-source data and provide accurate data support. The diagnosis module is based on convolutional neural networks and long-short-term memory networks to automatically identify potential faults. Combined with the visual result output module, it provides a scientific basis for equipment maintenance and ensures that the safety performance test results of the centrifugal releaser are more realistic and reliable. BRIEF DESCRIPTION OF THE DRAWINGS

[0013] The present invention will be further described in detail below with reference to the accompanying drawings and specific implementation methods.

[0014] Figure 1 Schematic diagram of the system principle; Figure 2 A flowchart of the steps. DETAILED DESCRIPTION

[0015] The present invention will be described in detail below with reference to the accompanying drawings and in combination with embodiments. It should be noted that, unless there is a conflict, the embodiments and features in the embodiments of the present application can be combined with each other.

[0016] According to one aspect of the present invention, Figure 1-Figure 2 As shown, a centrifugal releaser safety performance detection system is provided, including: a data acquisition module, which uses a multi-sensor array, including a MEMS acceleration sensor (such as ADI ADXL345, range ±16g, resolution 13 bits, sampling frequency 10kHz), a Hall-effect speed sensor (such as Allegro A1302, response frequency 0-100kHz, measurement error ±0.5%), and a piezoresistive pressure sensor (accuracy ±0.5% FS, range 0-10MPa). Each sensor is precisely triggered to acquire data through a time synchronization algorithm. The conversion relationship between the acceleration sensor output voltage and physical acceleration is:

[0017] in is the output voltage, is the reference voltage, is the measuring range, is the sensitivity (e.g. ADXL345 sensitivity is 3.9mg / LSB), is the supply voltage (unit V); The time synchronization algorithm is based on the Network Time Protocol (NTP) framework and combines the sliding window interpolation method to calibrate the timing of sensor data with different sampling frequencies. First, a unified reference timestamp is obtained through the NTP protocol. , and then the original acquisition time stamp of the sensor Make corrections, the correction formula is:

[0018] in is the local reference time of the sensor. At the same time, the linear interpolation formula is used to account for the time axis differences of multi-source data:

[0019] in Align time points for the target, and are adjacent original time points, and To correspond to the original data, the sensor data collected at different frequencies are resampled to a unified time scale to ensure accurate matching of multi-dimensional data in time series, laying the foundation for subsequent work.

[0020] The data preprocessing module performs preliminary processing on the collected data to remove noise and redundant information. This not only reduces the pressure on data transmission but also quickly extracts key data features, significantly improving data processing efficiency and reducing the computational burden of subsequent modules. Specifically, the preprocessing module integrates the Kalman filter algorithm and short-time Fourier transform (STFT) to denoise the raw data and extract time-frequency domain features, thereby achieving data dimensionality reduction. The Kalman filter algorithm constructs a state space model, uses the state of the previous moment to predict the current state, and then makes corrections based on the observed value. Its core formula is: State update equation:

[0021] Observation equation:

[0022] And the Kalman gain calculation:

[0023] in is the state vector, is the observed value, is the state transition matrix, is the observation matrix, To observe the noise covariance, we can effectively filter the data noise and make the data closer to the real state; At the same time, the short-time Fourier transform performs local analysis on the signal, and the formula is:

[0024] Convert time domain data to the time-frequency domain, extract key features such as speed fluctuation and vibration spectrum, complete data dimensionality reduction, and lay a solid foundation for efficient data processing in subsequent modules.

[0025] The data interaction module plays a bridging role, enabling smooth data flow between different modules, so that data can be transmitted to where it is needed in a timely and accurate manner, ensuring the coordinated work of various parts of the system and enhancing the integrity and coherence of the system; The module builds a wireless communication network based on IoT technology and uses the MQTT protocol (message queue telemetry transmission) to implement data interaction between edge computing devices, virtual simulation modules, physical simulation modules, and diagnostic modules. The publish / subscribe mechanism reduces network bandwidth usage, and data transmission delay can be controlled within 50ms. The transmission efficiency formula is:

[0026] Compression algorithms (such as LZ77) are used to increase the raw data compression ratio to 3:1, ensuring stable transmission of multi-source data in heterogeneous network environments and resolving the data silo problem of traditional detection systems. At the same time, the data interaction module integrates edge computing nodes (based on ARM architecture processors) and follows the principle of "local processing + cloud collaboration". Feature extraction and screening are performed locally on the collected raw data, and only key parameters (such as fault feature vectors and operating condition indicators) are uploaded to the cloud. The local processing logic of the edge node can be expressed as follows:

[0027] in It is a preprocessing function that includes operations such as Kalman filtering denoising and feature dimensionality reduction. It not only reduces the load on cloud servers, but also ensures the timeliness of data processing through the 10ms real-time response capability of edge nodes, avoiding network congestion and delay problems in the traditional cloud centralized processing mode.

[0028] The simulation module simulates the operating status of the centrifugal releaser under various actual working conditions and verifies it with real data. This makes up for the shortcomings of relying solely on actual testing or virtual simulation, greatly improving the authenticity and accuracy of the working condition simulation, so that the test results are more in line with the actual operation of the equipment. This module includes: Virtual simulation module, using computer simulation software ANSYS to build a 3D virtual model of the centrifugal releaser, inputting actual physical parameters and working conditions, and performing simulation analysis; The virtual simulation module is based on finite element analysis (FEA) and multi-body dynamics simulation technology, and constructs the structural mechanics equations of the centrifugal releaser:

[0029] in is the stiffness matrix, is the displacement vector, is the external force vector, and with the multibody dynamics equation:

[0030] in is the mass matrix, is the damping matrix, It is a generalized coordinate vector that enables accurate calculation of stress distribution and motion state under centrifugal force. For example, it can simulate the maximum stress of a spring leaf under centrifugal force through meshing (minimum element size 0.1mm). Physical simulation module: Using variable frequency motors, programmable controllers, and hydraulic systems, we build a physical test platform that can simulate different slopes, speeds, and loads to obtain operating data in real environments. Among them, the physical simulation module realizes PID closed-loop control through a programmable controller (PLC), and the control algorithm is:

[0031] in, is the control quantity output, is the systematic error (i.e. the difference between the target value and the actual value), 、 、 are the proportional, integral, and differential coefficients respectively.

[0032] The data fusion module deeply integrates pre-processed data with virtual physical simulation data, eliminating differences and contradictions between the data, unifying data formats and standards, and providing more accurate and reliable data support for subsequent analysis; Among them, the data fusion module adopts a dynamic weight allocation mechanism to adjust the fusion weight of sensor data and virtual simulation data in real time according to the working condition parameters. The weight calculation follows the formula:

[0033] in The confidence level of the data source is determined (e.g. the confidence level of physical sensors is determined by accuracy calibration, and the confidence level of virtual data is determined by simulation error). Multi-source data synchronization is achieved through the timestamp alignment algorithm, using the linear interpolation formula:

[0034] in Align time points for the target, and are adjacent original time points, and To correspond to the original data, the pre-processed data and virtual physical simulation data collected at different frequencies are resampled to ensure the accurate matching of multi-dimensional data in time series and feature dimensions, and to construct a high-quality fusion data set.

[0035] The diagnosis module builds a fault identification model based on convolutional neural networks (CNN) and long short-term memory networks (LSTM); Among them, this module extracts the time-frequency domain features of the vibration signal through one-dimensional convolution operation. The formula is:

[0036] in For the Tier The output of neurons, is the weight, is the bias, is the ReLU activation function, and the LSTM gating mechanism is combined to realize temporal feature learning. The forget gate formula is:

[0037] in is the Sigmoid function, is the hidden state at the previous moment, is the current input, and the fault classification is finally completed through the Softmax function:

[0038] in is the output of the fully connected layer, The number of fault categories is used to automatically identify faults such as bearing wear and spring failure.

[0039] The output module presents the test results in the form of visual charts and text descriptions based on the diagnostic report generated by the diagnostic module; Among them, this module converts the fault probability vector output by CNN+LSTM (such as the bearing wear probability of 0.92 and the normal state probability of 0.08) into a dynamic line chart, with the horizontal axis as the time series and the vertical axis as the fault confidence. At the same time, it generates a text report through the feature importance matrix (for example, the test shows that the spring leaf stress exceeds the threshold by 23%, and it is recommended to replace it within 48 hours).

[0040] Of course, the above description is not a limitation of the present invention, and the present invention is not limited to the above examples. Changes, modifications, additions or substitutions made by ordinary technicians in this technical field within the essential scope of the present invention also fall within the scope of protection of the present invention.

Claims

1. A centrifugal releaser safety performance detection system, characterized in that: The following steps are involved: S1: With the help of the data acquisition module, multiple types of sensors are used to comprehensively collect the operating data of the centrifugal releaser to ensure that all kinds of information during the operation of the equipment can be obtained, providing a rich and comprehensive data foundation for subsequent testing, and avoiding inaccurate test results due to missing data; S2: The preprocessing module performs preliminary processing on the collected data, denoising the raw data and extracting time-frequency domain features to achieve data dimensionality reduction and remove noise and redundant information. This not only reduces the pressure on data transmission but also quickly extracts key data features, significantly improving data processing efficiency and reducing the computational burden of subsequent modules. S3: The data interaction module acts as a bridge, enabling smooth data flow between different modules. This allows data to be transmitted to where it is needed in a timely and accurate manner, ensuring the coordinated operation of all parts of the system and enhancing the integrity and coherence of the system. S4: The simulation module simulates the operating status of the centrifugal releaser under various actual working conditions and verifies it with real data. This makes up for the shortcomings of relying solely on actual testing or virtual simulation, greatly improving the authenticity and accuracy of working condition simulation, so that the test results are more in line with the actual operation of the equipment; S5: The data fusion module deeply integrates the pre-processed virtual simulation data with the physical simulation data, eliminating differences and contradictions between the data, unifying the data format and standards, and providing more accurate and reliable data support for subsequent analysis; S6: The diagnostic module uses intelligent algorithms to analyze fused data, automatically identify potential equipment failures and generate diagnostic reports.

2. A centrifugal releaser safety performance detection system according to claim 1, characterized in that: The simulation module includes: a virtual simulation module and a physical simulation module; Virtual simulation module: Use computer simulation software ANSYS to build a three-dimensional virtual model of the centrifugal releaser, input actual physical parameters and working conditions, and perform simulation analysis; Physical simulation module: Using variable frequency motors, programmable controllers and hydraulic systems, we build a physical test platform that can simulate different slopes, speeds and loads to obtain operating data in real environments.

3. A centrifugal releaser safety performance detection system according to claim 1, characterized in that: The data acquisition module adopts a multi-sensor array, including a MEMS acceleration sensor, a Hall speed sensor and a piezoresistive pressure sensor. Each sensor realizes precise triggering of data acquisition through a time synchronization algorithm.

4. A centrifugal releaser safety performance detection system according to claim 1, characterized in that: The preprocessing module integrates the Kalman filter algorithm and the short-time Fourier transform to perform denoising and time-frequency domain feature extraction on the original data to achieve data dimensionality reduction.

5. A centrifugal releaser safety performance detection system according to claim 1, characterized in that: The data fusion module adopts a dynamic weight allocation mechanism to adjust the fusion weight of sensor data and virtual simulation data in real time according to working condition parameters, and realizes synchronous processing of multi-source data through a timestamp alignment algorithm.

6. A centrifugal releaser safety performance detection system according to claim 1, characterized in that: The diagnosis module builds a fault recognition model based on convolutional neural networks and long short-term memory networks.

7. A centrifugal releaser safety performance detection system according to claim 1, characterized in that: The result output module presents the test results in the form of visual charts and text descriptions based on the diagnosis report generated by the diagnosis module.

Images