A lighting fixture failure detection apparatus and method

Abstract

This invention discloses a lighting fixture fault detection device and method, relating to the field of lighting equipment operation and maintenance technology. It includes a front-end acquisition unit, a core processing unit, an edge computing unit, a fault feature database storage unit, a communication unit, an early warning execution unit, and a power supply unit. The front-end acquisition unit is electrically connected to the core processing unit and transmits the acquired operating parameters to it. The core processing unit is electrically connected to the edge computing unit, the fault feature database storage unit, the communication unit, and the early warning execution unit. The edge computing unit is electrically connected to the fault feature database storage unit. This invention, employing the aforementioned lighting fixture fault detection device and method, can accurately locate the fault type and location, has a fast fault response speed, strong anti-interference capability, and achieves predictive maintenance and safety protection through hierarchical management. It also supports model self-learning iteration, has strong adaptability, and significantly reduces the operation and maintenance costs of lighting systems.

Description

Technical Field

[0001] This invention relates to the field of lighting equipment operation and maintenance technology, and in particular to a lighting fixture fault detection device and detection method. Background Technology

[0002] With urban construction and industrial development, the application scenarios for lighting fixtures are constantly expanding. Large-scale lighting fixture clusters are deployed in municipal roads, tunnels, industrial plants, large commercial complexes, sports stadiums, and other scenarios. The long-term stable operation of lighting fixtures is directly related to public safety, production efficiency, and operating costs.

[0003] Currently, there are two main methods for fault detection of lighting fixtures: one is manual inspection, where maintenance personnel periodically check the operating status of the fixtures on-site. This method suffers from low inspection efficiency, high maintenance costs, delayed fault detection, and safety hazards associated with working at heights. It is especially difficult to conduct manual inspections in special scenarios such as tunnels and high-pole streetlights. The other method is a simple online detection method, which uses current and voltage sensors installed in the power supply circuit to detect the continuity of the circuit and determine whether the fixture is faulty. This method can only identify whether the fixture is completely malfunctioning, but it cannot accurately locate the type of fault (such as light source aging, driver power supply failure, poor line contact, harmonic abnormalities, etc.) or distinguish the location of the fault. It is also susceptible to power grid fluctuations, resulting in a high false alarm rate and a high missed alarm rate, and it cannot achieve early warning and hierarchical management of faults. Summary of the Invention

[0004] The purpose of this invention is to provide a lighting fixture fault detection device and method, which can accurately locate the fault type and location, has a fast fault response speed, strong anti-interference ability, and achieves predictive maintenance and safety protection through hierarchical management.

[0005] This invention provides a lighting fixture fault detection device, comprising a front-end acquisition unit, a core processing unit, an edge computing unit, a fault feature database storage unit, a communication unit, an early warning execution unit, and a power supply unit; the front-end acquisition unit is electrically connected to the core processing unit and transmits the acquired operating parameters to the core processing unit; the core processing unit is electrically connected to the edge computing unit, the fault feature database storage unit, the communication unit, and the early warning execution unit, and the edge computing unit is electrically connected to the fault feature database storage unit.

[0006] Preferably, the front-end acquisition unit includes an electrical parameter acquisition module, a temperature acquisition module, and an illumination acquisition module. The electrical parameter acquisition module includes a voltage acquisition submodule, a current acquisition submodule, and a harmonic analysis submodule. The temperature acquisition module includes a contact temperature sensor, which is attached to the lamp light source module, the driver power supply housing, and the power supply circuit terminals. The illumination acquisition module includes a lamp-end illumination sensor and an ambient illumination sensor.

[0007] Preferably, the early warning execution unit includes a local early warning module, a remote early warning module, and a loop protection module. The local early warning module includes an LED display screen and an audible and visual alarm, and the loop protection module includes a controllable relay.

[0008] Preferably, the communication unit includes a wired communication subunit and a wireless communication subunit; the wired communication subunit is an RS485 bus or an Ethernet communication module, and the wireless communication subunit is at least one of a LoRa communication module, a 4G / 5G communication module, and a WiFi communication module.

[0009] A method for detecting faults in lighting fixtures includes the following steps: Step S1: Benchmark calibration and fault feature library construction. Collect multi-dimensional benchmark operating parameters of different types of lighting fixtures under normal working conditions, simulate various typical fault scenarios to collect fault feature data, train a fault classification model based on machine learning algorithms, construct a standard fault feature library and store it in the fault feature library storage unit. Step S2: Synchronous acquisition of multi-dimensional operating parameters. Through the front-end acquisition unit, the real-time operating parameters of the target lighting fixture power supply circuit and the fixture body are synchronously acquired, and the acquired data is transmitted to the core processing unit. Step S3: Data preprocessing. The core processing unit performs noise reduction, filtering, and normalization on the received real-time operating parameters to eliminate acquisition errors caused by power grid fluctuations and environmental interference, and obtain standardized data to be detected. Step S4: Intelligent fault identification on the edge side. The edge computing unit calls the standard fault feature library and pre-trained fault classification model in the fault feature library storage unit to extract and match features of the standardized data to be detected, identify and determine whether the lamp has a fault. If a fault exists, the corresponding fault type, fault location and fault level result are output to the core processing unit. Step S5: Fault classification, early warning and protection execution. The core processing unit issues corresponding instructions to the early warning execution unit according to the fault level based on the received fault identification results. The early warning execution unit performs corresponding actions such as local audible and visual early warning, remote operation and maintenance push, and emergency disconnection of fault circuit.

[0010] Step S6: Data Upload and Model Iteration Optimization. The core processing unit uploads real-time operating data, fault identification records, and processing results to the cloud management platform through the communication unit. The cloud management platform optimizes and trains the fault classification model based on massive operation and maintenance data, and distributes the updated model and feature library to the fault feature library storage unit to complete the iterative upgrade.

[0011] A detection method for a lighting fixture fault detection device, in step S1, typical fault scenarios include light source short circuit / open circuit fault, light source aging and attenuation fault, drive power supply fault, power supply circuit short circuit / open circuit fault, poor contact of wiring terminals, and lamp overheating fault; the machine learning algorithm adopts random forest algorithm or lightweight CNN convolutional neural network algorithm to adapt to the computing power requirements of edge computing units.

[0012] Preferably, in step S3, the data preprocessing specifically involves first performing preliminary denoising on the collected real-time operating parameters using a moving average filtering algorithm, then eliminating power grid harmonic interference and environmental noise using a wavelet transform algorithm, and finally mapping the denoised data to the [0,1] interval using a min-max normalization algorithm to obtain standardized data to be detected.

[0013] Preferably, in step S4, the fault level is divided into three levels: Level 1 is an emergency fault, including short circuit in the power supply circuit, overheating of the lamp exceeding the safety threshold, and overheating of the wiring terminal, which poses a risk of fire and electrical safety; Level 2 is a major fault, including open circuit in the light source, failure of the drive power supply, and open circuit in the power supply circuit, which causes the lamp to be completely unable to light up; Level 3 is a general fault, including aging and decay of the light source, abnormal power factor, excessive harmonics, and current fluctuation faults caused by poor contact, in which the lamp can be lit normally but there is a risk of performance degradation or potential fault.

[0014] Preferably, in step S5, the specific rules for fault classification early warning and protection execution are as follows: For Level 1 emergency faults, the core processing unit immediately issues a cut-off command to the circuit protection module to disconnect the power supply to the faulty circuit, and simultaneously triggers the highest level alarm of the local audible and visual alarm, and pushes emergency fault alarm information to the cloud management platform and the mobile terminal of maintenance personnel through the remote early warning module; For Level 2 important faults, the core processing unit triggers a local audible and visual alarm, displays the fault type and location on the display screen, and simultaneously pushes the fault information to the cloud management platform through the remote early warning module, and generates and sends the corresponding maintenance work order to the mobile terminal of maintenance personnel; For Level 3 general faults, the core processing unit records and displays the fault information on the local display screen, and simultaneously uploads the fault data to the cloud management platform, incorporates it into the regular maintenance plan, and does not trigger audible and visual alarms.

[0015] Preferably, in step S2, multi-dimensional operating parameters are collected synchronously. Electrical parameters, temperature parameters, and illumination parameters are sampled synchronously using the same clock reference, with a sampling frequency of not less than 1kHz, to ensure the time synchronization of multi-dimensional parameters and improve the accuracy of fault feature identification.

[0016] Therefore, the present invention employs the above-mentioned lighting fixture fault detection device and detection method, which can accurately locate the fault type and location, has a fast fault response speed, strong anti-interference ability, and achieves predictive maintenance and safety protection of faults through hierarchical management and control.

[0017] The technical solution of the present invention will be further described in detail below with reference to the accompanying drawings and embodiments. Attached Figure Description

[0018] Figure 1 This is a schematic diagram of the overall structure of a lighting fixture fault detection device according to the present invention; Figure 2 This is a schematic diagram of the front-end acquisition unit in a lighting fixture fault detection device of the present invention; Figure 3 This is a schematic diagram of the early warning execution unit in a lighting fixture fault detection device of the present invention; Figure 4 This is a schematic diagram of the communication unit in a lighting fixture fault detection device of the present invention; Figure 5 This is a schematic diagram of the overall process of the detection method of the lighting fixture fault detection device of the present invention. Detailed Implementation

[0019] The technical solution of the present invention will be further described below with reference to the accompanying drawings and embodiments.

[0020] Unless otherwise defined, the technical or scientific terms used in this invention shall have the ordinary meaning as understood by one of ordinary skill in the art to which this invention pertains.

[0021] The terms "first," "second," and similar words used in this invention do not indicate any order, quantity, or importance, but are merely used to distinguish different components. Terms such as "comprising" or "including" mean that the element or object preceding the word encompasses the elements or objects listed after the word and their equivalents, without excluding other elements or objects. Terms such as "connected" or "linked" are not limited to physical or mechanical connections, but can include electrical connections, whether direct or indirect. Terms such as "upper," "lower," "left," and "right" are used only to indicate relative positional relationships; when the absolute position of the described object changes, the relative positional relationship may also change accordingly.

[0022] Example 1 like Figures 1-5 As shown, the present invention provides a lighting fixture fault detection device, comprising a front-end acquisition unit, a core processing unit, an edge computing unit, a fault feature database storage unit, a communication unit, an early warning execution unit, and a power supply unit.

[0023] The front-end acquisition unit is electrically connected to the core processing unit to synchronously acquire multi-dimensional operating parameters of the lighting fixture's power supply circuit and the fixture itself, and then transmits the acquired operating parameters to the core processing unit.

[0024] The core processing unit is electrically connected to the edge computing unit, the fault feature database storage unit, the communication unit, and the early warning execution unit. It is used to receive the operating parameters of the front-end acquisition unit, perform data preprocessing and forward it to the edge computing unit, and at the same time receive the fault identification results of the edge computing unit, drive the action of the early warning execution unit, and complete data interaction through the communication unit.

[0025] The edge computing unit is electrically connected to the fault feature library storage unit. It is used to call the pre-trained fault classification model and standard feature library in the fault feature library storage unit, perform fault feature matching and identification on the pre-processed operating parameters, and output the fault type, fault location and fault level results to the core processing unit.

[0026] The fault feature library storage unit is used to store the baseline parameter library of normal operating conditions of lighting fixtures, feature datasets of various fault types, and pre-trained fault classification models. It also supports receiving model update packages from the cloud to complete the iterative upgrade of the feature library and models.

[0027] The communication unit enables bidirectional data communication between the core processing unit and the cloud management platform and mobile maintenance terminal. The early warning execution unit executes corresponding levels of local early warning, remote early warning, and fault circuit protection actions based on instructions issued by the core processing unit. The power supply unit provides a stable power supply to all electrical modules within the device and is equipped with a backup power supply module for power outage scenarios. The front-end acquisition unit includes electrical parameter acquisition modules, temperature acquisition modules, and light intensity acquisition modules.

[0028] The electrical parameter acquisition module includes a voltage acquisition submodule, a current acquisition submodule, and a harmonic analysis submodule, which are used to acquire real-time voltage, real-time current, harmonic distortion rate, and power factor parameters of the lighting fixture power supply circuit, respectively. Specifically, the voltage acquisition submodule uses a precision voltage transformer, the current acquisition submodule uses a Hall current sensor, and the harmonic analysis submodule uses an ATT7053 energy metering chip, enabling the acquisition of voltage, current, power factor, and harmonic distortion rate with an accuracy of 0.5 class, and the sampling frequency is set to 2kHz.

[0029] The temperature acquisition module includes contact temperature sensors, which are mounted on the light source module, driver power supply housing, and power supply circuit terminals of the lamp to collect real-time temperature data of the core components and wiring locations of the lamp. The temperature acquisition module uses three sets of DS18B20 contact temperature sensors, respectively mounted on the aluminum substrate of the LED street light's light source module, driver power supply housing, and power supply circuit terminals, with a temperature measurement range of -55℃ to 125℃ and an accuracy of ±0.5℃. The illumination acquisition module uses two sets of BH1750 illumination sensors; one set is installed inside the light-emitting surface of the street light to collect the actual output illuminance, and the other set is installed on the outside of the lamp housing to collect ambient illuminance, with a measurement range of 0 to 65535 lx.

[0030] The illumination acquisition module includes a luminance sensor at the luminaire end and an ambient luminance sensor at the environment end, used to collect the actual output illuminance of the luminaire and the ambient illuminance of the luminaire installation environment, respectively. The early warning execution unit includes a local early warning module, a remote early warning module, and a loop protection module.

[0031] The local early warning module includes an LED display screen and an audible and visual alarm, used to display the operating status and fault information of the lighting fixtures locally, and to trigger audible and visual alarms when a fault occurs. The remote early warning module is used to push fault information, maintenance work orders, and early warning SMS messages to the cloud management platform and mobile maintenance terminals via a communication unit. The circuit protection module includes a controllable relay, connected in series to the power supply circuit of the lighting fixtures, used to cut off the power supply to the corresponding faulty circuit when an emergency fault command is received, to prevent the fault from escalating.

[0032] The communication unit includes a wired communication subunit and a wireless communication subunit; the wired communication subunit is an RS485 bus or an Ethernet communication module, and the wireless communication subunit is at least one of a LoRa communication module, a 4G / 5G communication module, and a WiFi communication module.

[0033] The core processing unit uses an STM32F407 microcontroller as the main control core, responsible for data reception, preprocessing, command issuance, and communication management. The edge computing unit uses an ESP32-S3 chip with a built-in AI accelerator, enabling local inference of lightweight CNN models to complete fault feature matching and identification. The fault feature library storage unit uses a W25Q128JV flash memory chip to store the benchmark parameter library, fault feature dataset, and pre-trained fault classification models.

[0034] The communication unit includes a LoRa communication module (SX1278) and a 4G communication module (EC200S). The LoRa module enables network communication between street light nodes, while the 4G module enables remote communication with the cloud management platform. The early warning execution unit includes an OLED display, an active buzzer and LED alarm lights forming an audible and visual alarm, and a 5A / 250V controllable relay connected in series to the street light's power supply circuit. The power supply unit uses an AC / DC switching power supply module with an input of AC220V and outputs of DC12V and DC3.3V to power each module. A 3.7V / 2000mAh lithium battery is also configured as a backup power source to support the device in reporting faults during grid outages.

[0035] The core processing unit communicates with the edge computing unit via the SPI bus. The core processing unit communicates with the front-end acquisition unit, the fault feature library storage unit, and the early warning execution unit via the I2C bus through the IO port. The communication unit is connected to the core processing unit via the UART serial port.

[0036] A method for detecting faults in lighting fixtures includes the following steps: Step S1: Benchmark calibration and fault feature database construction. Collect multi-dimensional benchmark operating parameters of different types of lighting fixtures under normal working conditions, including rated voltage AC220V, rated current 0.45A, power factor 0.95, total harmonic distortion (THD) ≤5%, normal operating temperature of light source substrate ≤60℃, temperature of drive power supply housing ≤55℃, and rated illuminance 12000lx.

[0037] Simulate various typical fault scenarios to collect fault characteristic data: Light source open circuit fault: current is 0, illuminance is 0, voltage is normal; Light source aging fault: current decays to below 70% of rated value, illuminance decays to below 60% of rated value, power factor decreases, harmonic distortion rate increases; Driver power supply fault: voltage is normal, current fluctuation exceeds ±20%, harmonic distortion rate THD ≥ 20%, power factor ≤ 0.7, driver power supply housing temperature rises abnormally; Power supply circuit short circuit fault: current instantaneously exceeds 5 times the rated value, voltage drops sharply; Terminal contact fault: current fluctuates periodically, terminal temperature exceeds 80℃; Lamp overheating fault: light source substrate temperature exceeds 85℃.

[0038] A fault classification model is trained based on machine learning algorithms, and a standard fault feature library is constructed and stored in the fault feature library storage unit. 1000 sets of valid data are collected for each fault scenario, and a lightweight CNN convolutional neural network algorithm is used to train the model, resulting in a fault classification model with a recognition accuracy of ≥98%. A standard fault feature library is then constructed and stored in the fault feature library storage unit.

[0039] In step S1, typical fault scenarios include light source short circuit / open circuit fault, light source aging and decay fault, drive power supply fault, power supply circuit short circuit / open circuit fault, poor contact of wiring terminals, and lamp overheating fault; the machine learning algorithm adopts random forest algorithm or lightweight CNN convolutional neural network algorithm to adapt to the computing power requirements of edge computing unit.

[0040] Step S2: Synchronous acquisition of multi-dimensional operating parameters. Through the front-end acquisition unit, the real-time operating parameters of the power supply circuit and the lamp body of the target lighting fixture are collected synchronously, and the collected data is transmitted to the core processing unit.

[0041] In step S2, multi-dimensional operating parameters are collected synchronously. Electrical parameters, temperature parameters, and illumination parameters are sampled synchronously using the same clock reference. The sampling frequency is not less than 1kHz to ensure the time synchronization of multi-dimensional parameters and improve the accuracy of fault feature identification.

[0042] The device is installed inside the control box of the LED street light pole. A relay is connected in series to the street light's power supply circuit, and all sensors in the front-end acquisition unit are installed as required. During operation, the core processing unit, using the same clock reference, controls the front-end acquisition unit to synchronously acquire the voltage, current, harmonic distortion rate, and power factor of the power supply circuit at a sampling frequency of 2kHz; the real-time temperature of the light source substrate, driver power supply, and terminals; and the luminous output and ambient illuminance of the lamp. This ensures that the time synchronization error of all acquired parameters is ≤1ms, and the acquired data is transmitted to the core processing unit in real time.

[0043] Step S3: Data preprocessing. The core processing unit performs noise reduction, filtering, and normalization on the received real-time operating parameters to eliminate acquisition errors caused by power grid fluctuations and environmental interference, and obtain standardized data to be tested.

[0044] In step S3, the data preprocessing specifically involves first using a moving average filtering algorithm to perform preliminary denoising on the collected real-time operating parameters, then using a wavelet transform algorithm to eliminate power grid harmonic interference and environmental noise, and finally using a min-max normalization algorithm to map the denoised data to the [0,1] interval to obtain standardized data to be detected.

[0045] Step S4: Intelligent fault identification on the edge side. The edge computing unit calls the standard fault feature library and the pre-trained CNN fault classification model in the fault feature library storage unit to extract and match features of the standardized data to be detected, identify and determine whether the lamp has a fault. If a fault exists, the corresponding fault type, fault location and fault level result are output to the core processing unit.

[0046] In step S4, the fault levels are divided into three levels: Level 1 is an emergency fault, including short circuit in the power supply circuit, overheating of the lamp (not less than 85°C), and terminal temperature (not less than 95°C), posing a fire and electrical safety risk. Level 2 is a critical fault, including open circuit in the light source, driver power supply failure, and open circuit in the power supply circuit, causing the lamp to be completely unable to light up. Level 3 is a general fault, including aging and attenuation of the light source, power factor not greater than 0.8, harmonic distortion rate not less than 15%, current fluctuation faults caused by poor contact, and terminal temperature not less than 70°C but less than 90°C; the lamp can light up normally but there is a risk of performance degradation or potential fault.

[0047] Step S5: Fault classification, early warning and protection execution. The core processing unit issues corresponding instructions to the early warning execution unit according to the fault level based on the received fault identification results. The early warning execution unit performs corresponding actions such as local audible and visual early warning, remote operation and maintenance push, and emergency disconnection of fault circuit.

[0048] In step S5, the specific rules for fault classification early warning and protection are as follows: For a level one emergency fault, the core processing unit immediately sends a cut-off command to the circuit protection module to disconnect the circuit power supply of the faulty street light, and at the same time triggers the highest level alarm of the audible and visual alarm. The emergency fault alarm information is pushed to the cloud management platform and the mobile terminal of the operation and maintenance personnel through the remote early warning module.

[0049] For Level 2 critical faults, the core processing unit triggers a low-frequency audible and visual alarm, displays the fault type and location on the screen, and simultaneously pushes the fault information to the cloud management platform through the remote alarm module, generating and issuing the corresponding maintenance work order to the mobile terminal of the maintenance personnel.

[0050] For Level 3 general faults, the core processing unit records and displays the fault information on the local display screen, and uploads the fault data to the cloud management platform for inclusion in the regular operation and maintenance plan, without triggering audible and visual alarms.

[0051] Step S6, Data Upload and Model Iteration Optimization: The core processing unit uploads real-time operating data, fault identification records and processing results to the cloud management platform through the communication unit. The cloud management platform optimizes and trains the fault classification model based on massive operation and maintenance data, and distributes the updated model and feature library to the fault feature library storage unit to complete the iterative upgrade.

[0052] Therefore, the present invention employs the above-mentioned lighting fixture fault detection device and detection method, which can accurately locate the fault type and location, has a fast fault response speed, strong anti-interference ability, realizes predictive maintenance and safety protection of faults through hierarchical management and control, supports model self-learning iteration, has strong adaptability, significantly reduces the operation and maintenance cost of lighting systems, improves operational safety and stability, and is suitable for fault detection and intelligent operation and maintenance of various lighting fixture clusters.

[0053] The above embodiments are only used to illustrate the technical solutions of the present invention and are not intended to limit them. Although the present invention has been described in detail with reference to preferred embodiments, those skilled in the art should understand that modifications or equivalent substitutions can still be made to the technical solutions of the present invention, and these modifications or equivalent substitutions cannot cause the modified technical solutions to deviate from the spirit and scope of the technical solutions of the present invention.

Claims

1. A lighting fixture fault detection device, characterized in that, It includes a front-end acquisition unit, a core processing unit, an edge computing unit, a fault feature database storage unit, a communication unit, an early warning execution unit, and a power supply unit; the front-end acquisition unit is electrically connected to the core processing unit and transmits the acquired operating parameters to the core processing unit; the core processing unit is electrically connected to the edge computing unit, the fault feature database storage unit, the communication unit, and the early warning execution unit, and the edge computing unit is electrically connected to the fault feature database storage unit.

2. A lighting fixture fault detection device according to claim 1, characterized in that, The front-end acquisition unit includes an electrical parameter acquisition module, a temperature acquisition module, and a light intensity acquisition module. The electrical parameter acquisition module includes a voltage acquisition submodule, a current acquisition submodule, and a harmonic analysis submodule. The temperature acquisition module includes a contact temperature sensor, which is attached to the lamp light source module, the driver power supply housing, and the power supply circuit terminals. The illumination acquisition module includes a lamp-end illumination sensor and an ambient illumination sensor.

3. A lighting fixture fault detection device according to claim 1, characterized in that, The early warning execution unit includes a local early warning module, a remote early warning module, and a loop protection module. The local early warning module includes an LED display screen and an audible and visual alarm, while the loop protection module includes a controllable relay.

4. A lighting fixture fault detection device according to claim 1, characterized in that, The communication unit includes a wired communication subunit and a wireless communication subunit; the wired communication subunit is an RS485 bus or Ethernet communication module, and the wireless communication subunit is at least one of a LoRa communication module, a 4G / 5G communication module, and a WiFi communication module.

5. The detection method of a lighting fixture fault detection device as described in any one of claims 1-4, characterized in that, Includes the following steps: Step S1: Benchmark calibration and fault feature library construction. Collect multi-dimensional benchmark operating parameters of different types of lighting fixtures under normal working conditions, simulate various typical fault scenarios to collect fault feature data, train a fault classification model based on machine learning algorithms, construct a standard fault feature library and store it in the fault feature library storage unit. Step S2: Synchronous acquisition of multi-dimensional operating parameters. Through the front-end acquisition unit, the real-time operating parameters of the target lighting fixture power supply circuit and the fixture body are synchronously acquired, and the acquired data is transmitted to the core processing unit. Step S3: Data preprocessing. The core processing unit performs noise reduction, filtering, and normalization on the received real-time operating parameters to eliminate acquisition errors caused by power grid fluctuations and environmental interference, and obtain standardized data to be detected. Step S4: Intelligent fault identification on the edge side. The edge computing unit calls the standard fault feature library and pre-trained fault classification model in the fault feature library storage unit to extract and match features of the standardized data to be detected, identify and determine whether the lamp has a fault. If a fault exists, the corresponding fault type, fault location and fault level result are output to the core processing unit. Step S5: Fault classification, early warning and protection execution. The core processing unit issues corresponding instructions to the early warning execution unit according to the fault level based on the received fault identification results. The early warning execution unit performs corresponding actions such as local audible and visual early warning, remote operation and maintenance push, and emergency disconnection of fault circuit. Step S6: Data Upload and Model Iteration Optimization. The core processing unit uploads real-time operating data, fault identification records, and processing results to the cloud management platform through the communication unit. The cloud management platform optimizes and trains the fault classification model based on massive operation and maintenance data, and distributes the updated model and feature library to the fault feature library storage unit to complete the iterative upgrade.

6. The detection method of a lighting fixture fault detection device according to claim 5, characterized in that, In step S1, typical fault scenarios include light source short circuit / open circuit fault, light source aging and decay fault, drive power supply fault, power supply circuit short circuit / open circuit fault, poor contact of wiring terminals, and lamp overheating fault. The machine learning algorithms employ random forest or lightweight CNN convolutional neural network algorithms to adapt to the computing power requirements of edge computing units.

7. The detection method of a lighting fixture fault detection device according to claim 5, characterized in that, In step S3, the data preprocessing specifically involves first using a moving average filtering algorithm to perform preliminary denoising on the collected real-time operating parameters, then using a wavelet transform algorithm to eliminate power grid harmonic interference and environmental noise, and finally using a min-max normalization algorithm to map the denoised data to the [0,1] interval to obtain standardized data to be detected.

8. The detection method of a lighting fixture fault detection device according to claim 5, characterized in that, In step S4, the fault level is divided into three levels: Level 1 is an emergency fault, including short circuit in the power supply circuit, overheating of the lamp exceeding the safety threshold, and overheating of the wiring terminal, which poses a fire and electrical safety risk; Level 2 is a critical fault, including open circuit in the light source, failure of the driver power supply, and open circuit in the power supply circuit, which causes the lamp to be completely unable to light up. Level 3 is a general fault, including light source aging and decay, abnormal power factor, excessive harmonics, and current fluctuation faults caused by poor contact. The lamp can be lit normally, but there is a risk of performance degradation or potential failure.

9. The detection method of a lighting fixture fault detection device according to claim 5, characterized in that, In step S5, the specific rules for fault classification warning and protection are as follows: For a level one emergency fault, the core processing unit immediately sends a cut-off command to the circuit protection module to disconnect the power supply to the fault circuit, and at the same time triggers the highest level alarm of the local audible and visual alarm, and pushes emergency fault alarm information to the cloud management platform and the mobile terminal of the operation and maintenance personnel through the remote warning module. For Level 2 critical faults, the core processing unit triggers a local audible and visual alarm, displays the fault type and location on the screen, and simultaneously pushes the fault information to the cloud management platform through the remote alarm module, generating and issuing the corresponding maintenance work order to the mobile terminal of the maintenance personnel. For Level 3 general faults, the core processing unit records and displays the fault information on the local display screen, and uploads the fault data to the cloud management platform for inclusion in the regular operation and maintenance plan, without triggering audible and visual alarms.

10. The detection method of a lighting fixture fault detection device according to claim 5, characterized in that, In step S2, multi-dimensional operating parameters are collected synchronously. Electrical parameters, temperature parameters, and illumination parameters are sampled synchronously using the same clock reference. The sampling frequency is not less than 1kHz to ensure the time synchronization of multi-dimensional parameters and improve the accuracy of fault feature identification.

Images