Real-time monitoring and dynamic analysis system supporting hot plug of multiple Bluetooth sensors

Abstract

The invention discloses a real-time monitoring and dynamic analysis system supporting hot plug of multiple Bluetooth sensors, which is used for solving the problems of unstable connection, incapability of dynamic maintenance and high-load data processing delay during cooperative monitoring of multiple wireless sensors. The system distributes an independent data channel for each three-axis acceleration Bluetooth sensor through a dynamic resource isolation module; disconnection and automatic reconnection are achieved through the connection state management module, and continuous monitoring is guaranteed; data acquisition and calculation analysis are decoupled through a logic decoupling processing assembly line module and a unified calculation resource scheduling module, and frequency domain and time domain feature extraction of multiple devices is executed in parallel based on a thread pool; through a dynamic visualization mapping module, dynamic binding and comparison display between any online sensor data and a display view in operation are realized. According to the invention, multi-device hot plug management is realized, continuous loss-free acquisition and efficient parallel analysis under a high sampling rate are realized, and the reliability, flexibility and real-time performance of a monitoring system are improved.

Description

Technical Field

[0001] This invention belongs to the field of industrial equipment condition monitoring technology, specifically relating to a real-time monitoring and dynamic analysis system that supports hot-swapping of multiple Bluetooth sensors, suitable for condition monitoring of rotating machinery. Background Technology

[0002] Condition monitoring of wind turbine gearboxes is a crucial aspect of ensuring the safe and stable operation of wind power equipment. Traditional monitoring solutions mainly fall into two categories: wired and wireless, both of which have significant drawbacks. Wired sensor deployment is complex, requiring cables to be laid inside the gearbox, leading to turbine downtime for hours or even days for modifications, resulting in high costs. While existing wireless sensor solutions avoid physical wiring, they face severe challenges in scenarios with complex electromagnetic environments, such as wind turbine towers. The dense metal structure inside the tower generates strong multipath interference and shielding effects on wireless signals such as Bluetooth, leading to highly unstable connections and frequent abnormal disconnections when multiple sensors are operating concurrently. Existing systems lack effective connection status maintenance and automatic recovery mechanisms, often requiring manual intervention to restart after a disconnection, severely disrupting the continuity of condition monitoring.

[0003] Furthermore, as monitoring demands evolve towards multi-point and high-precision monitoring, the deep-seated contradictions of traditional system architectures are becoming increasingly apparent. Their rigid equipment management logic cannot support the dynamic addition or deletion of sensors while the system is running; that is, it does not support hot-swapping, and adding a monitoring point requires a system shutdown and reset. At the data processing level, the system cannot efficiently coordinate the concurrent data streams generated by multiple high-sampling-rate sensors, often resulting in data loss or update delays due to uneven allocation of computing resources or link congestion. The mixed display of multi-channel data easily obscures key fault characteristic frequencies; during computationally intensive analyses such as spectrum and envelope analysis, the main thread is easily blocked, causing interface lag. These deficiencies not only affect the timely capture and accurate diagnosis of early, subtle fault characteristics in gearboxes but also hinder the construction of intelligent and predictive maintenance systems for wind farms. Therefore, there is an urgent need for a new monitoring system that can adapt to complex interference environments, support hot-swapping and connection self-recovery, and achieve reliable parallel processing and collaborative analysis of multi-device data. Summary of the Invention

[0004] To address the aforementioned technical problems, this invention provides a real-time monitoring and dynamic analysis system that supports hot-swapping of multiple Bluetooth sensors.

[0005] The technical solution of this invention to solve the above-mentioned technical problems is: a real-time monitoring and dynamic analysis system supporting hot-plugging of multiple Bluetooth sensors, the system comprising:

[0006] The dynamic resource isolation module is configured to: dynamically allocate and maintain a set of independent data receiving threads and multi-level buffer structures for each successfully connected triaxial acceleration Bluetooth sensor, so as to achieve physical isolation of data streams between different sensors, and automatically destroy all exclusive resources corresponding to the sensor when it is removed.

[0007] The connection status management module is configured to automatically add a connected sensor to the retry queue and perform a limited number of delayed reconnection attempts when an abnormal disconnection is detected, so as to achieve hot recovery of the connection.

[0008] The logic decoupling pipeline module, connected to the connection state management module, is configured to decouple the data sampling and receiving process from each sensor from the subsequent computationally intensive analysis process, so that the continuity of data sampling is not affected by the instantaneous fluctuations of the backend computing load.

[0009] The unified computing resource scheduling module, connected to the logical decoupled processing pipeline module, is configured to maintain a shared thread pool for receiving and executing parallel computing tasks from various sensors. The computing tasks include fast Fourier transform and time-domain feature extraction of peak-to-peak value, standard deviation, and skewness.

[0010] The dynamic visualization mapping module is configured to provide multiple independent display views and allow data streams from any online sensor to be dynamically bound or unbound to a specified display view during system operation to achieve real-time comparison display.

[0011] The aforementioned real-time monitoring and dynamic analysis system that supports hot-plugging of multiple Bluetooth sensors includes a multi-level buffer structure in the dynamic resource isolation module, comprising independent raw data buffers, X / Y / Z three-channel data buffers, and independent spectrum calculation buffers that are independent of each other and set sequentially according to the data processing stage.

[0012] The independent raw data buffer is configured as a first-in-first-out byte stream buffer to receive unparsed data frames from the Bluetooth communication link, in order to absorb wireless link jitter and avoid data loss due to upper-layer processing delays; the independent raw data buffer performs real-time frame structure verification on the received byte stream based on a preset data frame format, and discards data frames that fail the verification.

[0013] The X / Y / Z three-channel data buffer is configured as multiple independent double-ended queues corresponding one-to-one with the physical channels of the triaxial acceleration Bluetooth sensor. After the data frame passes the verification, its load is parsed into time-domain signal data of multiple physical channels, and stored separately after the range and sensitivity conversion is completed, so as to serve as the data source for time-domain waveform display and time-domain feature calculation.

[0014] The independent spectrum calculation buffer is configured to cache time-domain data blocks to be analyzed in the frequency domain and is continuously filled in a sliding update or cyclic overwrite manner to provide data input with consistent length and continuous time to the frequency domain analysis task, thereby supporting the stable execution of real-time spectrum calculation.

[0015] The aforementioned real-time monitoring and dynamic analysis system that supports hot-swapping of multiple Bluetooth sensors employs a progressive delay strategy when the connection status management module performs delayed reconnection attempts. It also maintains an independent retry counter for each device, stopping reconnection and triggering the device removal process after reaching the maximum number of attempts.

[0016] The aforementioned real-time monitoring and dynamic analysis system supporting hot-plugging of multiple Bluetooth sensors, in addition to decoupling the data sampling and receiving process from the subsequent analysis and processing process, is further configured as follows:

[0017] The system monitors the occupancy status of each data buffer structure in real time and dynamically adjusts the triggering timing or concurrency of computing tasks based on the occupancy level of the buffer structure. This reduces the concurrency intensity of computing tasks when the computing resource load is high or the buffer occupancy is close to the threshold, thereby ensuring the continuity of the data sampling process and the stability of the overall system operation.

[0018] The aforementioned real-time monitoring and dynamic analysis system that supports hot-plugging of multiple Bluetooth sensors is further configured with a unified computing resource scheduling module that dynamically adjusts the allocation of computing tasks and pauses or cancels the execution of computing tasks on channels in a hidden state based on the user interface's control signal for the display / hidden status of specific data channels.

[0019] The aforementioned real-time monitoring and dynamic analysis system that supports hot-swapping of multiple Bluetooth sensors includes a dynamic visualization mapping module that responds to user operations through an exclusive selection control in the user interface. When a user selects to bind a sensor to a view, an exclusive data channel between the sensor and the view is immediately established, and the data connection between the view and any previously bound sensors is automatically disconnected.

[0020] The aforementioned real-time monitoring and dynamic analysis system supporting hot-plugging of multiple Bluetooth sensors further includes a non-blocking advanced analysis and processing module configured as follows:

[0021] For the sensor currently selected by the dynamic visualization mapping module, a background analysis thread is started in a window independent of the system's main interface;

[0022] In the background analysis thread, a Hilbert transform is performed on the time-domain signal of the sensor to obtain the envelope signal and the envelope spectrum;

[0023] The analysis results are displayed in real time in the separate window, and interactive controls are provided for users to adjust the frequency range of the spectrum display.

[0024] The above-mentioned real-time monitoring and dynamic analysis system supporting hot-plugging of multiple Bluetooth sensors, wherein the real-time monitoring and dynamic analysis method of the system is characterized by comprising the following steps:

[0025] Step S1: Dynamic device access and resource isolation: In response to the Bluetooth sensor connection request, after the connection is successfully established, an independent data receiving thread and a multi-level buffer structure are dynamically created for the sensor.

[0026] Step S2: Connection status maintenance and hot recovery: During system operation, if an abnormal disconnection of a sensor connection is detected, it will be automatically added to the retry queue and a delayed reconnection attempt will be performed;

[0027] Step S3: Continuous high-sampling-rate data acquisition and verification: Through the independent data receiving thread and buffer structure, data is continuously received at a set sampling rate; frame structure verification is performed on each received data packet, invalid data packets are discarded, and the buffer structure absorbs the instantaneous delay caused by wireless link jitter to ensure that no valid data points are lost;

[0028] Step S4: Decoupling and Parallel Computing: The ready data collected and verified in Step S3 is encapsulated into computing tasks. The computing tasks include Fast Fourier Transform and calculation of peak-to-peak value, standard deviation and skewness. The data is submitted to a shared thread pool for parallel computing to achieve decoupling between the data acquisition process and the computing analysis process.

[0029] Step S5: Temperature data extraction and dynamic display: Extract temperature values ​​from the verified data frames, convert them, and dynamically update them to the specified position on the main interface according to the sensor access order;

[0030] Step S6: Visual dynamic mapping and display: Based on user interaction commands, establish or update the dynamic binding relationship between the sensor and the display view, and push the data or calculation results of the corresponding sensor to the bound view for display in real time;

[0031] Step S7: Device hot removal: In response to a sensor removal command or reconnection failure, sequentially terminate its data receiving thread, destroy its buffer structure, unbind its visual binding relationship, and release all associated resources;

[0032] Step S8: On-demand advanced analysis: In response to user requests, perform envelope analysis on the currently selected sensor in an independent analysis environment and present the results in a separate window.

[0033] The above-mentioned real-time monitoring and dynamic analysis system that supports hot-plugging of multiple Bluetooth sensors, in step S3, the system can achieve high-frequency data acquisition without data loss at 16kHz for a single device and 4kHz for each of two devices. The key is that: by establishing an independent buffer structure for each sensor, data competition between devices is isolated, and by offloading the computing task to the thread pool in step S4, the computing delay is avoided from blocking the data acquisition process.

[0034] The aforementioned real-time monitoring and dynamic analysis system supporting hot-plugging of multiple Bluetooth sensors further includes a data persistence step in its real-time monitoring and dynamic analysis method.

[0035] Create a separate data storage file for each connected sensor;

[0036] The raw data received and verified in step S3, the processing results generated in steps S4 and S5, and the corresponding timestamps are written into the corresponding files according to a predetermined format.

[0037] When performing thermal removal of the device in step S7, the device is safely shut down and the data file corresponding to the sensor is saved.

[0038] The beneficial effects of this invention are as follows:

[0039] 1. By integrating automatic reconnection mechanism for connection status management and a multi-level buffer structure with device-level isolation, the connection stability and data acquisition continuity of multiple wireless sensors in collaborative monitoring in complex industrial environments are improved, thereby reducing the risk of data loss caused by wireless link fluctuations.

[0040] 2. Enables flexible reconfiguration and efficient operation and maintenance of the monitoring system during operation. Thanks to the dynamic resource isolation and dynamic visualization mapping mechanism, users can dynamically connect, remove sensors, or switch display comparison objects without interrupting the overall system operation, realizing true hot-swappable devices and hot-binding views, greatly improving the convenience of system deployment, measurement point adjustment, and daily maintenance.

[0041] 3. It provides multi-level, high-efficiency parallel data processing and feature extraction capabilities. Through a logically decoupled processing pipeline and unified computing resource scheduling, the system can perform frequency domain transformation, envelope analysis, and multi-dimensional time domain feature calculations such as peak-to-peak value, standard deviation, and skewness on multiple devices in parallel. This not only fully taps the computing potential of the hardware and accelerates the analysis response speed, but also provides a rich feature information foundation for real-time assessment of device status and accurate diagnosis of early faults. Attached Figure Description

[0042] Figure 1 This is a block diagram of the overall architecture of the present invention.

[0043] Figure 2 This is a schematic diagram of the device access and initialization management process in an embodiment of the present invention.

[0044] Figure 3 This is a schematic diagram illustrating the detailed resource cleanup process for hot removal of equipment in an embodiment of the present invention.

[0045] Figure 4 This is a schematic diagram of the main process of the reconnection mechanism in an embodiment of the present invention.

[0046] Figure 5 This is a schematic diagram of the retry queue and retry count management details within the reconnection mechanism in this embodiment of the invention.

[0047] Figure 6 This is a schematic diagram of the independent multi-level buffer system established for each triaxial acceleration Bluetooth sensor device and its mapping relationship with the computing task in an embodiment of the present invention.

[0048] Figure 7 This is a schematic diagram illustrating how the unified computing resource scheduling module schedules and manages parallel computing tasks across multiple devices in an embodiment of the present invention.

[0049] Figure 8 This is a schematic diagram of the functional partitions and data flow of the human-computer interaction interface of the monitoring system in this embodiment of the invention.

[0050] Figure 9 This is a schematic diagram of the non-blocking advanced analysis and processing module performing envelope analysis and display in an embodiment of the present invention. Detailed Implementation

[0051] To make the objectives, technical solutions, and advantages of this invention clearer, the following description is provided in conjunction with the appendix. Figure 1-9 The present invention will be further described in detail below with reference to specific embodiments. It should be understood that the specific embodiments described herein are for illustrative purposes only and do not constitute a limitation thereof.

[0052] The overall system architecture and workflow are as follows: Figure 1 As shown in the figure, the real-time monitoring and dynamic analysis system that supports hot-plugging of multiple Bluetooth sensors provided by the embodiments of the present invention has the core design idea of ​​managing multiple triaxial acceleration Bluetooth sensor devices in a coordinated manner through a modular, decoupled and parallel software architecture.

[0053] The system's operating point is the Bluetooth communication interface, which is responsible for establishing wireless connections with physical sensors and transmitting and receiving data. The system includes:

[0054] Dynamic Resource Isolation Module: Whenever a new Bluetooth sensor successfully connects, this module dynamically allocates a completely independent set of computing and memory resources for it, including a dedicated data receiving thread and such... Figure 6The multi-level buffer structure shown provides physical isolation, ensuring that data streams from different sensors do not interfere with each other.

[0055] Connection status management module: When a connected sensor is detected to have disconnected abnormally, it automatically adds the sensor to the retry queue and performs a limited number of delayed reconnection attempts to achieve hot recovery of the connection;

[0056] Logical decoupling pipeline module: This module acts as a data flow relay, receiving raw or pre-processed data from the independent buffers of each sensor. Its core function is to decouple the continuous, high-speed data acquisition process from the computational analysis process, which may introduce delays. It encapsulates the data into standard, asynchronously executed computational task units.

[0057] The unified computing resource scheduling module maintains a globally shared thread pool. It receives various computing tasks submitted by the logically decoupled processing pipeline module and intelligently schedules and executes them in parallel based on the busy / idle status of the thread pool. These computing tasks specifically include Fast Fourier Transform to obtain spectral data, and extraction of peak-to-peak value, standard deviation, and skewness time-domain features.

[0058] The dynamic visualization mapping module manages the flexible correspondence between data and display. It provides multiple independent display views and allows users to dynamically bind real-time data streams or analysis results from any online sensor to selected views through interactive operations during system operation, thereby enabling multi-view waveform comparison display of data from multiple devices.

[0059] These modules work together through well-defined interfaces, forming a complete technology chain from reliable wireless data reception and parallel, efficient processing to flexible visualization of results.

[0060] The prerequisite for achieving reliable monitoring in this invention is the ability to manage the "online" status of the sensors. Figures 2 to 5 Together, they depicted a complete closed loop for equipment lifecycle management.

[0061] Figure 2 The process from device detection to stable monitoring is demonstrated. After the system starts, the user can trigger a scan, which lists nearby three-axis accelerometer Bluetooth sensor devices. The dynamic resource isolation module and the connection status management module work together in this process.

[0062] The system will query local records to check if the device address is a historical device address. If so, the corresponding checkbox will be automatically selected; otherwise, it will wait for the user to manually select it, aiming to achieve an automatic recovery experience after a disconnection. Whether the selection is automatic or manual, an asynchronous connection will be triggered, and the connection command will be sent via the Bluetooth interface.

[0063] If the connection is successful, the dynamic resource isolation module is immediately triggered to allocate an independent thread and buffer for the device. The specific process is as follows: Figure 6 As shown, the device then enters a stable monitoring state and begins continuous data interaction. If a connection fails, it immediately initiates connection failure handling and a reconnection mechanism. The system does not simply report an error, but rather... Figure 4 and Figure 5 The logic shown triggers an automatic reconnection process, demonstrating the system's resilience.

[0064] Figure 3 This document details the resource cleanup process for hot-removing a device. If the system is in a normal state during detection, it will continue to receive, process, and display data normally. However, when hot-removing a device from an other connection state, this process can be triggered by two events: the user actively unchecks the checkbox, or the system detects an abnormal disconnection and reconnects. The specific process is as follows: Figure 4 As shown.

[0065] When the user actively unchecks the box, the device hot removal process will be triggered immediately. The specific steps are as follows:

[0066] Stop data reception: Terminates the dedicated data reception thread for this device.

[0067] Disabling Bluetooth communication connection: releases underlying communication resources.

[0068] Destroy all dedicated buffers: This is the reverse operation of the dynamic resource isolation module when it is removed, cleaning up... Figure 6 All buffers allocated for this device, as shown, are released from memory.

[0069] Unbind the visual: The dynamic visual mapping module will unbind the device from any display views it may be bound to, and clear the views.

[0070] Release memory and object resources: Finally, clean up all software objects.

[0071] The entire removal process ensures that the operation of other devices is not affected, achieving complete isolation of resources and state.

[0072] Figure 4 and Figure 5 This section explains from different perspectives how the connection state management module achieves hot recovery of the connection. Figure 4 It is a description of the main logic.

[0073] When an abnormal disconnection is detected, the system first determines whether it was a user-initiated disconnection. If so, it proceeds to... Figure 3 If the normal removal process is not followed, the reconnection mechanism will be initiated, the system will add the device identifier to the retry queue, and start / reset the retry timer.

[0074] The retry timer is triggered periodically, driving the system to traverse the retry queue and initiate a reconnection for each device in the queue. If the reconnection is successful, the device returns to the online state and is removed from the queue, entering a stable monitoring state; if it fails, the device identifier is retained in the retry queue, waiting for the next attempt.

[0075] Figure 5 The retry strategy has been further refined by introducing a key control parameter: a maximum number of attempts threshold, which is set to 2 in this example. When a device sends a connection request, a connection status check is performed, and a reconnection attempt is made. If the connection is successful, the device enters a stable monitoring state; if it fails, the device is added to the retry queue, a retry timer is started, and the timer triggers to iterate through the retry queue.

[0076] Before each attempt, check the device's retries count. If the count is less than 2, reconnect and increment the retries count by 1, regardless of success or failure.

[0077] If reconnection fails, the device remains in the queue. If the number of attempts is ≥2, the system abandons the device, deeming it a final failure, and subsequently triggers... Figure 3 The device hot removal process shown combines a gradual delay strategy with a limited number of attempts to distinguish between transient interference and permanent failures, thus maintaining the connection as much as possible while avoiding the unnecessary waste of system resources.

[0078] The specific implementation methods of the elastic buffer system and parallel computing scheduling are as follows: Figure 6 , Figure 7 As shown, in order to achieve continuous and lossless data acquisition and efficient analysis, this invention constructs an isolated buffer system for each device and processes high-concurrency data streams through a decoupled parallel architecture.

[0079] Figure 6 The diagram clearly demonstrates the independent data processing pipeline built by the dynamic resource isolation module for each successfully connected sensor. This multi-level buffer structure is key to the system's high reliability and performance, and specifically includes three levels:

[0080] Independent Raw Data Buffer: This is a first-in, first-out (FIFO) byte stream buffer used to receive raw, unparsed data packets from the Bluetooth link. Its primary function is to absorb wireless link jitter and prevent packets from being discarded by the underlying driver due to insufficient data processing time. The buffer is managed based on a preset data frame format, performing real-time verification on the received byte stream and discarding invalid packets that do not conform to the format.

[0081] X / Y / Z three-channel data buffer: After the original data frame passes the verification, its load is parsed into acceleration digital signals of three physical channels, and after range and sensitivity conversion, it is stored in three independent double-ended queues. These queues constitute the direct data source for the system to perform time-domain waveform display and time-domain feature calculation.

[0082] Independent spectrum calculation buffer: This is usually a fixed-length array used to cache time-domain data blocks to be analyzed in the frequency domain. New data is continuously added and updated in a sliding window or circular overlay manner to ensure the real-time performance of spectrum calculation.

[0083] These three buffer layers correspond to different subsequent computational tasks: real-time waveform analysis directly uses the channel buffer data for plotting; the eigenvalue and envelope calculation task obtains data from the channel buffer; and the real-time waveform spectrum calculation task obtains data blocks from the spectrum calculation buffer for FFT transformation.

[0084] When the sensor is removed from the device Figure 6 The buffer destruction unit at the bottom will automatically clean up the corresponding third-level buffer of the removed device, realizing the immediate reclamation of resources.

[0085] Figure 7 This reveals how the logically decoupled processing pipeline module and the unified computing resource scheduling module work together to achieve efficient parallel processing. After the data from each sensor buffer is ready, the logically decoupled processing pipeline module encapsulates it into standardized asynchronous computing tasks, which are then submitted to a global thread pool managed by the unified computing resource scheduling module.

[0086] The core responsibility of this module is to unify the scheduling of computing resources. As a central scheduler, it receives and manages various computing tasks:

[0087] Spectrum calculation task: Perform a Fast Fourier Transform on the spectrum calculation buffer data of the device and output spectrum data.

[0088] Peak-to-peak value, standard deviation, and skewness calculation task: Calculate the time-domain statistical characteristic values ​​of the latest segment of data in the device channel buffer.

[0089] Envelope and envelope spectrum calculation tasks: triggered by advanced analysis and executed in a background thread.

[0090] All these tasks are executed in parallel under the management of a global thread pool. The size of the thread pool is configurable, which effectively utilizes the computing power of multi-core CPUs and avoids single-threaded processing bottlenecks.

[0091] The calculation results are passed back to the main thread through a thread-safe signal / message mechanism, which ultimately drives the interface update and completes the real-time display of waveforms and feature values.

[0092] This design completely decouples the I / O-intensive process of data acquisition from the computationally intensive process of data analysis, ensuring that the continuity of data acquisition is never affected by the backend computing load, even at high sampling rates and with multiple devices running concurrently.

[0093] The specific process of implementing dynamic visualization mapping and advanced analysis is as follows: Figure 8 as well as Figure 9 As shown.

[0094] Figure 8 This is a schematic diagram of the functional areas and data flow of the human-computer interaction interface of the monitoring system. The interface is mainly divided into four functional areas:

[0095] Device Management Area: Displays all discovered and connected devices in a list format. Users can perform device management operations such as scanning, connecting (checking), and disconnecting (unchecking) in this area.

[0096] Waveform Display Area: This is the core display area of ​​the dynamic visualization mapping module. Considering the clarity of the display interface and waveform comparison, the settings interface can display a maximum of two waveforms and spectrum diagrams from connected Bluetooth devices. Each view can independently and completely display the three-channel waveforms (X / Y / Z) and their spectrum diagram of a bound device. The checkbox above the user view selects the device to be displayed. During system operation, data streams from any online sensor are dynamically bound to the specified view.

[0097] When a new device is bound to a view, the system immediately establishes an exclusive data channel and automatically disconnects any devices previously bound to that view, enabling instantaneous and clear switching of view content and supporting real-time comparison display.

[0098] Feature Display Area: This area displays the peak-to-peak values, standard deviation, and skewness of the time-domain feature values ​​calculated by all connected devices, as well as the real-time monitored temperature values.

[0099] Envelope Analysis: Provides an envelope analysis button. The availability of this button and the analysis target are usually associated with the device currently active (bound) in the waveform display area, displaying the device's envelope and envelope spectrum.

[0100] Figure 9 The workflow of the non-blocking advanced analysis processing module is described in detail, demonstrating how complex Hilbert transforms and envelope spectrum analyses can be executed in an independent environment, avoiding blocking the main thread.

[0101] The process begins when the user clicks the envelope analysis button on the interface. The system first determines the target device, which is either a single device or a dual device currently bound to the waveform display view through the dynamic visualization mapping module.

[0102] Next, a separate window is created and displayed for the corresponding device. Simultaneously, the system acquires the latest X / Y / Z three-channel data (from...) of the target device. Figure 6 (The data is obtained from the channel data buffer shown). To avoid interfering with the main thread, the system creates a background analysis thread to pass the data in.

[0103] In the background analysis thread, the core calculation steps are executed sequentially:

[0104] Perform Hilbert transform: Apply Hilbert transform to the time-domain signal of each channel to generate the corresponding device envelope.

[0105] Performing an FFT on the envelope signal yields the envelope spectrum: Performing an FFT on the calculated envelope signal yields its frequency domain representation, i.e., the envelope spectrum.

[0106] After the calculation is complete, the result is sent to the UI main thread via a thread-safe signal-slot mechanism.

[0107] After receiving the results, the main thread passes them to the created window. This window is responsible for rendering the results of multi-device collaborative display, typically showing the envelope waveforms and envelope spectra of different devices in a paginated or side-by-side manner. More importantly, this window implements a real-time update mechanism, continuously receiving new calculation results from background threads and refreshing the display, while also providing interactive controls for users to adjust the spectrum display range.

[0108] The entire analysis process is conducted independently outside the main monitoring interface, completely isolating complex calculations and ensuring an extremely smooth master data acquisition and display process.

[0109] In summary, this invention, through the specific embodiments described above, constructs a complete technical solution with interconnected and collaborative components. From the device-level resource independence ensured by the dynamic resource isolation module, to the connection resilience guaranteed by the connection status management module; from the efficient parallel processing achieved by the logic decoupling processing pipeline module and the unified computing resource scheduling module, to the flexible observation methods provided by the dynamic visualization mapping module; and finally to the non-blocking advanced analysis processing module for handling complex tasks, each part is innovatively designed to address the pain points pointed out in the background technology. The combined effect of these modules ultimately enables continuous, reliable, real-time acquisition, processing, and intelligent analysis of high-sampling-rate data in complex industrial scenarios with dynamic access to multiple Bluetooth sensors, significantly improving the overall performance and practicality of the monitoring system.

[0110] The above description is merely a preferred embodiment of the present invention. Although embodiments of the present invention have been shown and described above, it is understood that the above embodiments are exemplary and should not be construed as limiting the present invention. Those skilled in the art can make changes, modifications, substitutions, and variations to the above embodiments within the scope of the present invention.

Claims

1. A real-time monitoring and dynamic analysis system supporting hot-swapping of multiple Bluetooth sensors, characterized in that, include: The dynamic resource isolation module is configured to: dynamically allocate and maintain a set of independent data receiving threads and multi-level buffer structures for each successfully connected triaxial acceleration Bluetooth sensor, so as to achieve physical isolation of data streams between different sensors, and automatically destroy all exclusive resources corresponding to the sensor when it is removed. The connection status management module is configured to automatically add a connected sensor to the retry queue and perform a limited number of delayed reconnection attempts when an abnormal disconnection is detected, so as to achieve hot recovery of the connection. The logic decoupling pipeline module, connected to the connection state management module, is configured to decouple the data sampling and receiving process from each sensor from the subsequent computationally intensive analysis process, so that the continuity of data sampling is not affected by the instantaneous fluctuations of the backend computing load. The unified computing resource scheduling module, connected to the logical decoupled processing pipeline module, is configured to maintain a shared thread pool for receiving and executing parallel computing tasks from various sensors. The computing tasks include fast Fourier transform and time-domain feature extraction of peak-to-peak value, standard deviation, and skewness. The dynamic visualization mapping module is configured to provide multiple independent display views and allow data streams from any online sensor to be dynamically bound or unbound to a specified display view during system operation to achieve real-time comparison display.

2. The system according to claim 1, characterized in that, The multi-level buffer structure in the dynamic resource isolation module includes independent raw data buffers, X / Y / Z three-channel data buffers, and independent spectrum calculation buffers that are independent of each other and set sequentially according to the data processing stage. The independent raw data buffer is configured as a first-in-first-out byte stream buffer to receive unparsed data frames from the Bluetooth communication link, in order to absorb wireless link jitter and avoid data loss due to upper-layer processing delays. The independent raw data buffer performs real-time frame structure verification on the received byte stream based on a preset data frame format, and discards data frames that fail the verification. The X / Y / Z three-channel data buffer is configured as multiple independent double-ended queues corresponding one-to-one with the physical channels of the triaxial acceleration Bluetooth sensor. After the data frame passes the verification, its load is parsed into time-domain signal data of multiple physical channels, and stored separately after the range and sensitivity conversion is completed, so as to serve as the data source for time-domain waveform display and time-domain feature calculation. The independent spectrum calculation buffer is configured to cache time-domain data blocks to be analyzed in the frequency domain and is continuously filled in a sliding update or cyclic overwrite manner to provide data input with consistent length and continuous time to the frequency domain analysis task, thereby supporting the stable execution of real-time spectrum calculation.

3. The system according to claim 1, characterized in that, When the connection state management module performs a delayed reconnection attempt, it adopts a progressive delay strategy and maintains an independent retry counter for each device. After reaching the maximum number of attempts, it stops reconnecting and triggers the device removal process.

4. The system according to claim 1, characterized in that, The logic decoupling pipeline module, in addition to decoupling the data sampling and receiving process from the subsequent analysis and processing process, is further configured as follows: The system monitors the occupancy status of each data buffer structure in real time and dynamically adjusts the triggering timing or concurrency of computing tasks based on the occupancy level of the buffer structure. This reduces the concurrency intensity of computing tasks when the computing resource load is high or the buffer occupancy is close to the threshold, thereby ensuring the continuity of the data sampling process and the stability of the overall system operation.

5. The system according to claim 1, characterized in that, The unified computing resource scheduling module is also configured to: dynamically adjust the allocation of computing tasks and suspend or cancel the execution of computing tasks on channels in a hidden state based on the display / hidden status control signal of a specific data channel in the user interface.

6. The system according to claim 1, characterized in that, The dynamic visualization mapping module responds to user operations through an exclusive selection control in the user interface; when the user selects to bind a sensor to a view, an exclusive data channel from the sensor to the view is immediately established, and the data connection between the view and any previously bound sensors is automatically disconnected.

7. The system according to any one of claims 1 to 6, characterized in that, The system also includes a non-blocking advanced analytics module configured as follows: For the sensor currently selected by the dynamic visualization mapping module, a background analysis thread is started in a window independent of the system's main interface; In the background analysis thread, a Hilbert transform is performed on the time-domain signal of the sensor to obtain the envelope signal and the envelope spectrum; The analysis results are displayed in real time in the separate window, and interactive controls are provided for users to adjust the frequency range of the spectrum display.

8. A method for real-time monitoring and dynamic analysis based on the system according to any one of claims 1 to 7, characterized in that, Includes the following steps: Step S1: Dynamic device access and resource isolation: In response to the Bluetooth sensor connection request, after the connection is successfully established, an independent data receiving thread and a multi-level buffer structure are dynamically created for the sensor. Step S2: Connection status maintenance and hot recovery: During system operation, if an abnormal disconnection of a sensor connection is detected, it will be automatically added to the retry queue and a delayed reconnection attempt will be performed; Step S3: Continuous high-sampling-rate data acquisition and verification: Through the independent data receiving thread and buffer structure, data is continuously received at a set sampling rate; frame structure verification is performed on each received data packet, invalid data packets are discarded, and the buffer structure absorbs the instantaneous delay caused by wireless link jitter to ensure that no valid data points are lost; Step S4: Decoupling and Parallel Computing: The ready data collected and verified in Step S3 is encapsulated into computing tasks. The computing tasks include Fast Fourier Transform and calculation of peak-to-peak value, standard deviation and skewness. The data is submitted to a shared thread pool for parallel computing to achieve decoupling between the data acquisition process and the computing analysis process. Step S5: Temperature data extraction and dynamic display: Extract temperature values ​​from the verified data frames, convert them, and dynamically update them to the specified position on the main interface according to the sensor access order; Step S6: Visual dynamic mapping and display: Based on user interaction commands, establish or update the dynamic binding relationship between the sensor and the display view, and push the data or calculation results of the corresponding sensor to the bound view for display in real time; Step S7: Device hot removal: In response to a sensor removal command or reconnection failure, sequentially terminate its data receiving thread, destroy its buffer structure, unbind its visual binding relationship, and release all associated resources; Step S8: On-demand advanced analysis: In response to user requests, perform envelope analysis on the currently selected sensor in an independent analysis environment and present the results in a separate window.

9. The method according to claim 8, characterized in that, In step S3, the system can achieve high-frequency data acquisition without data loss at 16kHz for a single device and 4kHz for each of two devices. The key is that the independent buffer structure established for each sensor isolates data competition between devices. At the same time, in step S4, the computing task is offloaded to the thread pool, avoiding computing delays from blocking the data acquisition process.

10. The method according to claim 8, characterized in that, The method also includes a data persistence step: Create a separate data storage file for each connected sensor; The raw data received and verified in step S3, the processing results generated in steps S4 and S5, and the corresponding timestamps are written into the corresponding files according to a predetermined format. When performing thermal removal of the device in step S7, the device is safely shut down and the data file corresponding to the sensor is saved.

Images