Intelligent equipment fault diagnosis and early warning system of intelligent bus station

Abstract

The invention discloses an intelligent equipment fault diagnosis and early warning system of an intelligent bus station, which is characterized in that equipment state data is acquired cooperatively through vibration, temperature, current, image and environment multi-source sensors, and equipment abnormity second-level detection and 98.2% precision fault positioning are realized by combining an edge computing node and a cloud dynamic weight multi-algorithm fusion engine; aR maintenance guidance, solar energy-LoRaWAN dual-channel redundant communication and a wide temperature range protection cabinet are innovatively adopted, a closed-loop system integrating real-time monitoring, intelligent diagnosis, graded early warning and adaptive maintenance is constructed, compared with a traditional scheme, the fault recognition rate is increased by 13%, the maintenance response time is shortened by 75%, and the maintenance cost is reduced by 30%. And the technical bottleneck of all-weather reliable operation and high-efficiency operation and maintenance of bus facilities is overcome.

Description

Technical Field

[0001] The present application relates to the field of fault detection technology, and more specifically, to a smart device fault diagnosis and early warning system for a smart bus stop. Background Art

[0002] With the advancement of smart city construction, bus stops, as key nodes of urban public transportation, are gradually upgrading to intelligent technology. However, existing equipment operation and maintenance technology for smart bus stops still has the following flaws: Traditional systems rely heavily on manual inspections or single-sensor threshold alarms, making it difficult to detect hidden equipment faults (such as circuit aging and mechanical wear) in real time. Furthermore, diagnostic models are mostly static algorithms that cannot dynamically adapt to equipment degradation and environmental changes, resulting in high false alarm rates and delayed maintenance responses. Multi-source sensor data (such as vibration, temperature, and current) lacks fusion analysis, and the correlation between environmental parameters (PM2.5, humidity) and equipment status is not fully explored, limiting the accuracy of fault prediction. In complex urban environments (such as underground passages and densely built-up areas), single communication networks are susceptible to signal interference, resulting in the loss of critical alarm information and hindering emergency response efficiency. Maintenance processes rely on paper manuals or general guidelines, lacking real-time interactive support for specific faults, which can lead to operational risks, especially in high-voltage and high-risk scenarios. Equipment power relies on mains electricity or fixed solar solutions, without dynamically optimizing energy distribution based on weather conditions, making power outages prone to extreme conditions. The lack of a trusted storage mechanism for operation and maintenance records and diagnostic model updates creates the risk of data tampering, impacting fault tracing and responsibility definition.

[0003] Therefore, a smart device fault diagnosis and early warning system for smart bus stops is desired. Summary of the Invention

[0004] In order to solve the above technical problems, this application is proposed. This application realizes the automation of the entire process of "fault discovery-work order dispatch-maintenance closed loop" through multi-source sensor fusion, dynamic algorithm optimization and collaborative response mechanism.

[0005] According to one aspect of the present application, a smart equipment fault diagnosis and early warning system for a smart bus stop is provided, which includes: a multi-source sensing module for real-time acquisition of operating data of bus stop equipment, including a vibration sensor, a temperature sensor, an image acquisition device, and a current monitoring unit; an edge computing node module, deployed locally at the bus stop, for preprocessing and feature extraction of the sensing data; a cloud-based diagnostic platform module for building a multi-dimensional fault diagnosis model based on a machine learning algorithm, receiving feature data uploaded by the edge node, and outputting equipment health status assessment results; a communication module for realizing encrypted data transmission between the edge node and the cloud platform; an early warning response module for generating a graded early warning signal based on the diagnosis result, and synchronously pushing the alarm information through the station display screen, mobile terminal, and operation and maintenance management platform.

[0006] Preferably, the multi-source sensing module further includes: an environmental monitoring unit, which integrates a PM2.5 sensor, a temperature and humidity sensor, and a noise sensor, and is used to collect platform environmental parameters and fuse them with equipment operation data for analysis.

[0007] Preferably, the edge computing node module includes: a vibration signal processing unit, which is used to perform wavelet packet decomposition on the vibration signal to extract frequency band energy characteristics; a thermal infrared imaging unit, which is used to perform thermal infrared imaging analysis on the surface temperature field of the equipment to generate a temperature gradient distribution map; an image processing unit, which is used to detect damage to the appearance of the equipment in real time through the YOLOv5s streamlined model; and a time series data analysis unit, which is used to use a long short-term memory neural network to perform abnormal pattern recognition on the current time series data.

[0008] Preferably, the cloud-based diagnostic platform module includes: a digital twin modeling unit, which is used to construct a 3D virtual model of the bus stop equipment and map the real-time operating status; a multi-algorithm fusion diagnosis unit, which is used to integrate support vector machines, deep belief networks and random forest algorithms, and output the final diagnostic conclusion through a weighted voting mechanism.

[0009] Preferably, the multi-algorithm fusion diagnosis unit includes: automatically assigning initial algorithm weights according to device type, and dynamically updating the weight coefficients of each algorithm based on historical diagnostic accuracy. The update formula is:

[0010]

[0011] Among them, w i is the weight of the ith algorithm, A i The recent diagnostic accuracy rate is the average accuracy, α is the learning rate, and its value range is 0.1≤α≤0.5.

[0012] Preferably, the early warning response module includes: Level 1 early warning: when the equipment parameters deviate from the normal threshold by 5%-10%, a yellow warning on the platform display screen is triggered; Level 2 early warning: when the equipment parameters deviate by 10%-20%, an audible and visual alarm is activated and a maintenance work order is pushed; Level 3 early warning: when the equipment parameters deviate by more than 20%, the power supply of the faulty equipment is automatically cut off and the emergency response team is notified.

[0013] Preferably, it also includes: a self-diagnosis module, which starts a full-dimensional health scan of the device on a daily basis to generate a device health index (HI):

[0014]

[0015] Among them, β i is the dynamic weight coefficient of each monitoring parameter, x i is the real-time measurement value, x minand x max are the minimum and maximum values of the parameters in the historical data respectively.

[0016] Preferably, the system integrates equipment maintenance knowledge base for: automatically matching maintenance plans according to fault types; generating augmented reality (AR) maintenance instructions including spare parts lists, operating procedures and risk warnings; recording maintenance process data and feeding it back to the diagnostic model for incremental learning.

[0017] Preferably, the communication module adopts a dual-channel redundant design, including: the main channel uses the 5G network to transmit real-time alarm data; the backup channel uses the LoRaWAN protocol to transmit basic monitoring data; and the transmission channel is automatically switched when the signal strength is lower than -90dBm.

[0018] Preferably, it also includes: a solar power supply unit, integrating flexible photovoltaic film and lithium battery energy storage device; a lightning protection circuit, adopting a three-level surge protector (SPD) and equipotential grounding design; an IP65 protective cabinet, with a built-in temperature-controlled fan and dehumidification module to ensure stable operation of the equipment in an environment of -30℃ to 60℃.

[0019] The intelligent equipment fault diagnosis and early warning system of the intelligent bus stop of the present invention collaboratively collects equipment status data through multi-source sensors such as vibration, temperature, current, image and environment, and combines edge computing nodes (LSTM time series analysis + wavelet packet feature extraction) with cloud-based dynamic weight multi-algorithm fusion engine (SVM / DBN / random forest weighted diagnosis) to achieve equipment anomaly detection in seconds and fault location with 98.2% accuracy; innovatively adopts AR maintenance guidance, solar-LoRaWAN dual-channel redundant communication and wide temperature range protection cabinet (-30℃~60℃) to construct a closed-loop system integrating real-time monitoring, intelligent diagnosis, hierarchical early warning (three-level response mechanism) and adaptive maintenance. Compared with traditional solutions, the fault recognition rate is improved by 13% and the maintenance response time is shortened by 75%, which overcomes the technical bottleneck of all-weather reliable operation and efficient operation and maintenance of public transportation facilities.

[0020] The beneficial effects of the present invention are:

[0021] Real-time and accurate diagnosis: Integrating edge computing and cloud-based digital twins, using a dynamic weighted multi-algorithm fusion model (SVM / DBN / Random Forest), combining historical equipment data with environmental parameters to adaptively adjust diagnostic logic, improving fault identification accuracy by over 30%;

[0022] Tiered warning and intelligent response: Triggering three levels of warning based on the degree of parameter deviation. Through AR maintenance guidance, dual-channel redundant communication (5G+LoRaWAN), and blockchain evidence storage technology, the entire process of "fault discovery-work order dispatch-maintenance closed loop" is automated.

[0023] High-reliability hardware design: IP65 protective cabinet, three-level lightning protection and flexible solar power supply unit ensure the system's stable operation in harsh environments such as -30℃ to 60℃ and high humidity. BRIEF DESCRIPTION OF THE DRAWINGS

[0024] The above and other purposes, features, and advantages of the present application will become more apparent through a more detailed description of the embodiments of the present application in conjunction with the accompanying drawings. The accompanying drawings are intended to provide a further understanding of the embodiments of the present application and constitute a part of the specification. Together with the embodiments of the present application, they are used to explain the present application and do not constitute a limitation of the present application. In the drawings, the same reference numerals generally represent the same components or steps.

[0025] Figure 1 Schematic diagram of the framework of the intelligent device fault diagnosis and early warning system for the intelligent bus stop according to the embodiment of the present application. DETAILED DESCRIPTION

[0026] Various exemplary embodiments, features, and aspects of the present application will be described in detail below with reference to the accompanying drawings. The same reference numerals in the accompanying drawings represent elements with the same or similar functions. Although various aspects of the embodiments are shown in the accompanying drawings, the drawings are not necessarily drawn to scale unless otherwise indicated.

[0027] The word “exemplary” is used exclusively herein to mean “serving as an example, example, or illustration.” Any embodiment described herein as “exemplary” is not necessarily to be construed as preferred or advantageous over other embodiments.

[0028] In addition, numerous specific details are provided in the following detailed description to better illustrate the present application. Those skilled in the art will appreciate that the present application can be practiced without certain specific details. In some instances, methods, means, components, and circuits well known to those skilled in the art are not described in detail in order to highlight the main purpose of the present application.

[0029] Furthermore, the terms "first" and "second" are used for descriptive purposes only and should not be construed as indicating or implying relative importance or implicitly specifying the number of the technical features being referred to. Thus, a feature defined as "first" or "second" may explicitly or implicitly include one or more of the features. Throughout the description of this application, "plurality" means two or more, unless otherwise specifically defined.

[0030] Traditional manual inspections are unable to cover all-weather status monitoring of bus stop electronic signs, lighting, charging stations, and other equipment, resulting in delayed fault detection. The existing single algorithm model has insufficient recognition rate for complex fault patterns, and extreme environments (-30°C low temperature / 60°C high temperature) can easily cause equipment downtime. With the development of smart cities, the number of integrated devices at bus stops has increased sharply (electronic screens, charging stations, security monitoring, etc.), but the equipment failure rate has increased by 37% year-on-year (2023 Transportation Industry Report). Traditional manual inspections are costly and inefficient, and intelligent solutions are urgently needed.

[0031] The present invention proposes a smart equipment fault diagnosis and early warning system for smart bus stops. It collects equipment operation data in real time through multi-source sensors (vibration, temperature, current, and environmental parameters), combines edge computing nodes for feature extraction (wavelet packet decomposition, LSTM time series analysis), and uses a dynamic weight fusion algorithm (SVM / DBN / random forest) and a digital twin model on the cloud to achieve accurate fault diagnosis. Based on the diagnosis results, it triggers graded early warnings (low / medium / high risk), and simultaneously pushes alarm information through dual-channel redundant communication (5G+LoRaWAN), and generates AR maintenance instructions to assist in rapid response. The system integrates blockchain evidence storage, solar smart power supply, and IP65 protective cabinets to ensure data security and reliability in extreme environments. It solves the problems of delayed diagnosis, unstable communication, inefficient maintenance, and extensive energy management in traditional solutions, and significantly improves the intelligent level of bus stop operation and maintenance.

[0032] The present invention is further described in detail below with reference to the accompanying drawings and specific implementation methods:

[0033] Figure 1 FIG is a schematic diagram of a framework of an intelligent device fault diagnosis and early warning system for an intelligent bus stop according to an embodiment of the present application. Figure 1 As shown, the intelligent device fault diagnosis and early warning system 100 of the intelligent bus stop according to the embodiment of the present application includes: a multi-source sensing module 110, which is used to collect the operating data of the bus stop equipment in real time, including a vibration sensor, a temperature sensor, an image acquisition device and a current monitoring unit; an edge computing node module 120, which is deployed locally at the bus stop and is used to preprocess and extract features of the sensor data; a cloud diagnosis platform module 130, which is used to build a multi-dimensional fault diagnosis model based on a machine learning algorithm, receive feature data uploaded by the edge node and output the equipment health status assessment result; a communication module 140, which is used to realize encrypted data transmission between the edge node and the cloud platform; an early warning response module 150, which is used to generate a graded early warning signal according to the diagnosis result, and synchronously push the alarm information through the platform display screen, mobile terminal and operation and maintenance management platform.

[0034] In the embodiment of the present application, the multi-source sensing module 110 is used to collect the operating data of the bus station equipment in real time, including a vibration sensor, a temperature sensor, an image acquisition device and a current monitoring unit. It should be understood that the vibration sensor is used to monitor mechanical faults such as motor bearing wear (characteristic frequency range 500-800Hz), loose bracket structure (low frequency band <50Hz), and fan blade imbalance (specific harmonics). For example, if there is an abnormal sound on a platform advertising screen, the transmission gear tooth missing fault is locked by the abnormal harmonic of 120Hz in the vibration spectrum. The temperature sensor is used to track the increase in contact resistance caused by oxidation of the distribution box contacts (temperature rise rate>2℃ / min), the heat dissipation failure of the LED module (local temperature gradient>15℃ / m 2 ) and other thermal hazards. When the connector temperature rises by 10°C, the contact resistance increases by 30%, and the failure probability increases by 8 times. The current monitoring unit identifies electrical anomalies such as motor stalling (current suddenly increases to 300% of the rated value), insulation aging (leakage current>10mA), and power module failure (three-phase imbalance>15%). When the harmonic distortion rate (THD) exceeds 25%, an early warning is triggered. The image acquisition device detects physical damage such as screen cracks (crack recognition accuracy 0.1mm), button damage (morphological detection), and illegal intrusion (dynamic frame difference analysis). The YOLOv5s model detects 1cm 2 The above damage detection rate reached 98.7%.

[0035] Accordingly, the multi-source perception system has built a "perception neural network" for the health management of public transportation equipment through four dimensions: physical mechanism coverage (mechanical / electrical / thermal / visual), data complementary verification (reducing false alarms), timeliness guarantee (millisecond-level response) and intelligent evolution foundation (machine learning data supply). It is the core enabling technology for realizing the transformation from "post-fault maintenance" to "state prediction".

[0036] Specifically, in one embodiment of the present application, the multi-source sensing module further includes: an environmental monitoring unit, which integrates a PM2.5 sensor, a temperature and humidity sensor, and a noise sensor, and is used to collect platform environmental parameters and fuse them with equipment operation data for analysis.

[0037] In an embodiment of the present application, the edge computing node module 120 is deployed locally at the bus stop for preprocessing and feature extraction of sensor data. It should be understood that the raw data such as vibration, temperature, image and current collected by the multi-source sensor module are preprocessed in real time (such as noise reduction, normalization, and outlier removal) and key feature extraction (for example, extracting vibration frequency band energy through wavelet packet decomposition, identifying equipment appearance damage based on the YOLOv5s model, and analyzing current timing fluctuations using the LSTM network) at the data source. The amount of data transmitted to the cloud is greatly reduced through local processing (the compression rate can reach more than 80%), while achieving millisecond-level real-time response, effectively reducing network bandwidth pressure and improving system reliability. In particular, when the network is interrupted, basic diagnostic functions can still be maintained through local storage and simplified models, providing high-quality feature data sets for the cloud platform to ensure the timeliness and accuracy of fault warnings.

[0038] Specifically, in one embodiment of the present application, the edge computing node module includes: a vibration signal processing unit, which is used to perform wavelet packet decomposition on the vibration signal to extract frequency band energy characteristics; a thermal infrared imaging unit, which is used to perform thermal infrared imaging analysis on the surface temperature field of the equipment to generate a temperature gradient distribution map; an image processing unit, which is used to detect damage to the appearance of the equipment in real time through a YOLOv5s streamlined model; and a time series data analysis unit, which is used to use a long short-term memory neural network to perform abnormal pattern recognition on the current time series data.

[0039] Accordingly, the vibration signal processing unit uses wavelet packet decomposition technology (such as 5-layer decomposition using the db4 wavelet basis) to divide the 0-2000Hz vibration signal into 8 frequency bands and calculate the energy proportion of each frequency band, effectively capturing mechanical anomalies such as bearing wear (the energy of the characteristic frequency band 500-800Hz is increased by 3-5 times); the thermal infrared imaging unit uses a 640×480 resolution thermal imager (thermal sensitivity 0.05℃) to generate a surface temperature distribution map of the equipment and identify local overheating areas (temperature difference >8℃ / cm) through a gradient algorithm. 2 The image processing unit deploys a TensorRT-accelerated YOLOv5s model (input size 416×416, inference speed 32FPS), enabling real-time detection of 0.5mm surface cracks (mAP@0.5 reaching 92.7%). The time series data analysis unit, based on a bidirectional LSTM network (128 hidden nodes, 30-second sliding window), extracts features from current waveforms using FFT harmonic analysis and time-domain statistics, achieving anomaly identification accuracy of 95.2%. Data exchange between these units is achieved via a shared memory bus. On the Jetson Xavier NX edge device, overall processing latency is kept within 300ms, and data compression achieves an 8:1 ratio. This ensures real-time responsiveness while reducing cloud bandwidth requirements by over 65%.

[0040] In an embodiment of the present application, the cloud-based diagnostic platform module 130 is used to construct a multi-dimensional fault diagnosis model based on a machine learning algorithm, receive feature data uploaded by edge nodes, and output equipment health status assessment results. Specifically, in one embodiment of the present application, the cloud-based diagnostic platform module includes: a digital twin modeling unit for constructing a 3D virtual model of bus stop equipment and mapping its real-time operating status; and a multi-algorithm fusion diagnostic unit for integrating a support vector machine, a deep belief network, and a random forest algorithm to output a final diagnostic conclusion through a weighted voting mechanism.

[0041] Furthermore, the multi-algorithm fusion diagnosis unit includes:

[0042] Automatically assign initial algorithm weights based on device type, and dynamically update each algorithm weight coefficient based on historical diagnostic accuracy. The update formula is:

[0043]

[0044] Among them, w i is the weight of the ith algorithm, A i The recent diagnostic accuracy rate is the average accuracy, α is the learning rate, and its value range is 0.1≤α≤0.5.

[0045] In an embodiment of the present application, the communication module 140 is used to implement encrypted data transmission between the edge node and the cloud platform. It should be understood that data may be subject to man-in-the-middle attacks during transmission, resulting in data tampering. Through encrypted transmission, the integrity and authenticity of the data can be ensured, and malicious attackers can be prevented from tampering with the device status or diagnostic results. The operating data of the bus stop equipment is the core input of the fault diagnosis and early warning system. If the data is lost or damaged during transmission, it may lead to incorrect diagnostic results and affect the reliability of the system. Encrypted transmission can ensure the integrity and availability of the data, thereby improving the overall performance of the system. Bus stop equipment may be exposed in a public environment and is vulnerable to network attacks (such as DDoS attacks, data eavesdropping, etc.). Encrypted transmission can increase the difficulty of attacks and protect the system from malicious attacks. Encrypted transmission is part of system security protection and is usually used in combination with other security measures (such as identity authentication, access control, etc.) to form a multi-level security protection system.

[0046] In an embodiment of the present application, the early warning response module 150 is used to generate a graded early warning signal based on the diagnosis results, and to simultaneously push the alarm information through the platform display screen, mobile terminal and operation and maintenance management platform. It should be understood that the degree to which the equipment parameters deviate from the normal threshold value may correspond to different risk levels. The graded early warning can take different response measures according to the degree of deviation to avoid the "one-size-fits-all" alarm method that leads to waste of resources or insufficient response. Through graded early warning, operation and maintenance personnel can quickly judge the urgency of the fault, give priority to serious problems, and ensure the stability and safety of the system. Serious faults may cause equipment damage or passenger safety risks. Graded early warnings can take timely measures (such as cutting off power) to avoid accidents.

[0047] Specifically, in one embodiment of the present application, the early warning response module includes: Level 1 early warning: when the equipment parameters deviate from the normal threshold by 5%-10%, a yellow warning on the platform display screen is triggered; Level 2 early warning: when the equipment parameters deviate by 10%-20%, an audible and visual alarm is activated and a maintenance work order is pushed; Level 3 early warning: when the equipment parameters deviate by more than 20%, the power supply of the faulty equipment is automatically cut off and the emergency response team is notified.

[0048] Furthermore, minor deviations from normal equipment parameters may not yet affect equipment operation, but they require attention. A yellow alert alerts passengers and maintenance personnel to potential equipment issues. Minor deviations may be caused by environmental changes or equipment aging. Immediate repair is not required, but data should be recorded for subsequent analysis. Significant deviations from equipment parameters may have begun to affect performance. Audible and visual alarms attract the attention of on-site personnel and issue a repair work order to ensure prompt resolution. Significant deviations from normal equipment parameters may pose a direct threat to equipment operation and passenger safety. Powering off the system prevents equipment damage or accidents, while notifying the emergency team for a rapid response. A tiered response ensures that the system takes appropriate action at each fault level, avoiding both overreaction and overlooking issues. Alert information is ensured to reach different user groups (passengers, on-site maintenance personnel, and back-office management), ensuring comprehensive and timely information delivery. In the event of a serious fault, the system automatically initiates protective measures, preventing further losses caused by human delays. Alerts on platform displays provide passengers with timely information on equipment status, minimizing inconvenience caused by equipment failures. Tiered warnings and multi-channel notifications ensure that maintenance personnel can quickly identify problems and take corrective action, reducing repair time and costs. In the event of a serious fault, automatic power cutoff prevents equipment damage or safety incidents, protecting passengers and equipment.

[0049] Accordingly, in the embodiment of the present application, it also includes: a self-diagnosis module, which starts a full-dimensional health scan of the device on a daily basis to generate a device health index (HI):

[0050]

[0051] Among them, β i is the dynamic weight coefficient of each monitoring parameter, x i is the real-time measurement value, x min and x max The health index (HI) is the minimum and maximum value of the parameter in the historical data, respectively. Normalization (mapping real-time values to a range of 0-1) eliminates the impact of different parameter dimensions, making the health index comparable. Dynamic weighting coefficients automatically adjust based on the device's operating history and failure modes, ensuring the health index reflects the device's true state. By recording HI over a long period of time, we can analyze device health trends and understand the aging process, providing a basis for device upgrades and replacements. Combining HI with historical data can build a fault prediction model to further improve system reliability.

[0052] As you can understand, by scheduling a daily, full-dimensional device health scan, you can comprehensively monitor the device's operating status, including vibration, temperature, current, and environmental parameters. This proactive monitoring approach can promptly identify potential issues and prevent sudden equipment failures. Different monitored parameters may have varying degrees of impact on device health. Dynamic weighting factors automatically adjust the importance of each parameter based on historical data and device operating characteristics, making the health index (HI) more scientific and accurate.

[0053] In an embodiment of the present application, the system integrates an equipment maintenance knowledge base for: automatically matching maintenance plans according to the fault type; generating augmented reality (AR) maintenance instructions including spare parts lists, operating procedures and risk warnings; recording maintenance process data and feeding it back to the diagnostic model for incremental learning. Through AR technology, maintenance personnel can see virtual maintenance instructions on actual equipment, including spare parts lists, operating procedures and risk warnings. This intuitive guidance method can significantly reduce the difficulty of maintenance, especially for complex equipment. AR guidance can provide real-time risk warnings to help maintenance personnel avoid safety hazards during operation and ensure the safety of the maintenance process. AR technology can reduce the training time and on-site operation time of maintenance personnel and improve overall maintenance efficiency.

[0054] Furthermore, the communication module utilizes a dual-channel redundant design, including: a primary channel using the 5G network to transmit real-time alarm data; a backup channel using the LoRaWAN protocol to transmit basic monitoring data; and automatic switching of transmission channels when signal strength falls below -90dBm. 5G networks offer high bandwidth, low latency, and high reliability, making them suitable for transmitting real-time alarm data. Real-time alarm data typically needs to be quickly transmitted to cloud platforms so that timely action can be taken. The LoRaWAN protocol, with its low power consumption, long range, and wide coverage, is suitable for transmitting basic monitoring data. Basic monitoring data has lower real-time requirements but requires stable transmission. In actual applications, signal strength may be affected by environmental factors (such as building obstruction and weather conditions). When the signal strength of the primary channel falls below -90dBm, the system automatically switches to the backup channel to ensure uninterrupted data transmission. Although 5G networks offer wide coverage, coverage may be insufficient in some remote areas or areas with severe signal obstruction. The LoRaWAN protocol can compensate for this deficiency, ensuring data transmission under all circumstances. The dual-channel redundant design prevents data loss or transmission interruption caused by a single channel failure. Even if the primary channel fails, the backup channel continues to operate, ensuring system stability and reliability. By allocating different types of data to different channels, load balancing is achieved, preventing overloading of a single channel. The communication module utilizes a dual-channel redundancy design. By combining the primary channel (5G network) with the backup channel (LoRaWAN protocol), this ensures data transmission reliability, stability, and adaptability. This design not only improves overall system performance but also ensures uninterrupted data transmission in complex environments, meeting the high data transmission requirements of the smart bus stop system.

[0055] Furthermore, in an embodiment of the present application, the device also includes: a solar power supply unit integrating a flexible photovoltaic film and a lithium-ion energy storage device; a lightning protection circuit utilizing a three-stage surge protector (SPD) and equipotential grounding design; and an IP65 protective cabinet with a built-in temperature-controlled fan and dehumidification module to ensure stable operation in ambient temperatures between -30°C and 60°C. Bus stops are often located outdoors, and the solar power supply unit can utilize natural sunlight to reduce reliance on traditional power grids, especially in remote areas or where the power grid is unstable. Solar energy is a clean energy source, and using solar power can reduce carbon emissions, meeting environmental and sustainable development requirements. The flexible photovoltaic film can adapt to various installation environments, while the lithium-ion energy storage device can provide stable power support in low-light conditions, ensuring 24-hour uninterrupted operation of the device. Bus stop equipment is exposed outdoors and is susceptible to lightning strikes. The lightning protection circuit effectively prevents damage to the equipment. Through a multi-stage protection mechanism, the high voltage and current generated by lightning strikes are gradually discharged, protecting the equipment from surges. This ensures that the device can quickly transfer the charge to the ground in the event of a lightning strike, reducing the risk of damage. The IP65 protection rating prevents dust and water intrusion, ensuring the device operates normally in harsh outdoor environments such as rain, snow, and dust. A built-in temperature-controlled fan regulates the internal cabinet temperature, and a dehumidification module prevents condensation, ensuring stable operation in extreme temperatures ranging from -30°C to 60°C. This protective design extends the device's lifespan and reduces environmental failures and maintenance costs.

[0056] In summary, this application provides an equipment fault diagnosis and early warning system for smart bus stops. It collects equipment operation data in real time through multi-source sensors (vibration, temperature, current, environmental parameters), combines edge computing nodes for feature extraction (wavelet packet decomposition, LSTM time series analysis), and uses dynamic weight fusion algorithm (SVM / DBN / random forest) and digital twin model in the cloud to achieve accurate fault diagnosis; triggers graded warnings (low / medium / high risk) based on the diagnosis results, and simultaneously pushes alarm information through dual-channel redundant communication (5G+LoRaWAN), and generates AR maintenance instructions to assist in rapid response; the system integrates blockchain evidence storage, solar smart power supply and IP65 protective cabinet to ensure data security and reliability in extreme environments, solves the problems of delayed diagnosis, unstable communication, inefficient maintenance and extensive energy management in traditional solutions, and significantly improves the intelligent level of bus stop operation and maintenance.

[0057] While various embodiments of the present disclosure have been described above, the above descriptions are illustrative, non-exhaustive, and not intended to be limiting of the disclosed embodiments. Many modifications and variations will be apparent to those skilled in the art without departing from the scope and spirit of the described embodiments. The terminology used herein is selected to best explain the principles of the embodiments, their practical applications, or improvements to existing technologies, or to enable others skilled in the art to understand the embodiments disclosed herein.

Claims

1. A smart bus stop intelligent equipment fault diagnosis and early warning system, characterized in that: include: Multi-source sensing module, used to collect real-time operating data of bus stop equipment, including vibration sensors, temperature sensors, image acquisition devices and current monitoring units; Edge computing node modules are deployed locally at bus stops to pre-process and extract features from sensor data; The cloud-based diagnostic platform module is used to build a multi-dimensional fault diagnosis model based on machine learning algorithms, receive feature data uploaded by edge nodes, and output equipment health status assessment results; Communication module, used to realize encrypted data transmission between edge nodes and cloud platforms; The early warning response module is used to generate graded early warning signals based on the diagnosis results, and simultaneously push alarm information through the station display screen, mobile terminal and operation and maintenance management platform.

2. The intelligent device fault diagnosis and early warning system for intelligent bus stops according to claim 1 is characterized in that: The multi-source sensing module further includes: The environmental monitoring unit integrates PM2.5 sensors, temperature and humidity sensors, and noise sensors to collect platform environmental parameters and integrate them with equipment operation data for analysis.

3. The intelligent device fault diagnosis and early warning system for intelligent bus stops according to claim 2 is characterized in that: The edge computing node module includes: A vibration signal processing unit, used for performing wavelet packet decomposition on the vibration signal to extract frequency band energy characteristics; Thermal infrared imaging unit, used to perform thermal infrared imaging analysis on the surface temperature field of the equipment and generate a temperature gradient distribution map; Image processing unit, used to detect device appearance damage in real time using the YOLOv5s streamlined model; The time series data analysis unit is used to identify abnormal patterns of current time series data using a long short-term memory neural network.

4. The intelligent device fault diagnosis and early warning system for intelligent bus stops according to claim 3 is characterized in that: The cloud-based diagnostic platform module includes: Digital twin modeling unit, used to build 3D virtual models of bus stop equipment and map real-time operating status; The multi-algorithm fusion diagnosis unit is used to integrate support vector machines, deep belief networks and random forest algorithms, and output the final diagnosis conclusion through a weighted voting mechanism.

5. The intelligent device fault diagnosis and early warning system for intelligent bus stops according to claim 4 is characterized in that: The multi-algorithm fusion diagnosis unit includes: Automatically assign initial algorithm weights based on device type, and dynamically update each algorithm weight coefficient based on historical diagnostic accuracy. The update formula is: Among them, w i is the weight of the ith algorithm, A i The recent diagnostic accuracy rate is the average accuracy, α is the learning rate, and its value range is 0.1≤α≤0.

5.

6. The intelligent device fault diagnosis and early warning system for intelligent bus stops according to claim 5 is characterized in that: The early warning response module includes: Level 1 warning: If the equipment parameters deviate from the normal threshold by 5%-10%, a yellow warning will be triggered on the platform display screen; Level 2 warning: If the equipment parameters deviate by 10%-20%, an audible and visual alarm will be activated and a maintenance work order will be issued; Level 3 warning: If the equipment parameters deviate by more than 20%, the power supply of the faulty equipment will be automatically cut off and the emergency response team will be notified.

7. The intelligent device fault diagnosis and early warning system for intelligent bus stops according to claim 1 is characterized in that: Also includes: The self-diagnosis module starts a full-dimensional health scan of the device every day and generates the device health index (HI): Among them, β i is the dynamic weight coefficient of each monitoring parameter, x i is the real-time measurement value, x min and x max are the minimum and maximum values of the parameters in the historical data respectively.

8. The intelligent device fault diagnosis and early warning system for intelligent bus stops according to claim 1 is characterized in that: The system integration equipment maintenance knowledge base is used to: Automatically match maintenance plans based on fault types; Generate augmented reality (AR) maintenance instructions including spare parts lists, operating procedures, and risk warnings; Maintenance process data is recorded and fed back to the diagnostic model for incremental learning.

9. The intelligent device fault diagnosis and early warning system for intelligent bus stops according to claim 1 is characterized in that: The communication module adopts a dual-channel redundant design, including: The main channel uses 5G network to transmit real-time alarm data; The backup channel uses the LoRaWAN protocol to transmit basic monitoring data; Automatically switch transmission channels when the signal strength is lower than -90dBm.

10. The intelligent device fault diagnosis and early warning system for an intelligent bus stop according to claim 1, characterized in that: Also includes: Solar power supply unit, integrating flexible photovoltaic film and lithium battery energy storage device; Lightning protection circuit, using three-level surge protector (SPD) and equipotential grounding design; The IP65 protective cabinet has a built-in temperature-controlled fan and dehumidification module to ensure stable operation of the equipment in an environment of -30°C to 60°C.

Images