Detection method and system of track lighting equipment, terminal and storage medium

Abstract

The invention discloses a track lighting equipment detection method and system, a terminal and a storage medium, and the method comprises the steps: obtaining a plurality of pieces of historical equipment data of a plurality of pieces of track lighting equipment, and carrying out the preprocessing, and obtaining the preprocessing equipment data; determining a preset neural network model, and performing model training and model fine tuning on the preset neural network model according to the preprocessed equipment data to obtain a target equipment health degree detection model; obtaining current equipment data of target track lighting equipment, and performing equipment health degree detection processing on the current equipment data through a target equipment health degree detection model to obtain a current equipment health value of the target track lighting equipment; and carrying out equipment processing strategy matching according to the current equipment health value to obtain a target equipment processing strategy, and carrying out equipment processing on the target track lighting equipment according to the target equipment processing strategy. According to the invention, the reliability of the track lighting equipment is effectively improved, and the normal operation of the track traffic is ensured.

Description

Technical Field

[0001] The present invention relates to the technical field of track lighting, and in particular to a detection method, system, terminal and computer-readable storage medium for track lighting equipment. Background Art

[0002] With the rapid development of urban rail transit, the safety, reliability and efficiency of rail transit lighting equipment are receiving more and more attention.

[0003] However, in the prior art, the inspection of rail transit lighting equipment generally relies mainly on regular manual inspections. However, the regular manual inspection method has high labor costs and incomplete data collection, which leads to poor health inspection results of rail transit lighting equipment and the safety of rail transit lighting equipment cannot be guaranteed.

[0004] Therefore, the prior art still needs to be improved and developed. Summary of the invention

[0005] The main purpose of the present invention is to provide a detection method, system, terminal and computer-readable storage medium for track lighting equipment, aiming to solve the problem that the detection of rail transit lighting equipment in the prior art generally relies mainly on manual periodic inspections, which has high labor costs and incomplete data collection, resulting in poor health detection effects of track lighting equipment and the safety of rail transit lighting equipment cannot be guaranteed.

[0006] To achieve the above object, the present invention provides a detection method for track lighting equipment, the detection method for track lighting equipment comprising the following steps: Acquire multiple historical device data of multiple track lighting devices, and pre-process the multiple historical device data to obtain pre-processed device data; Determine a preset neural network model, and perform model training and model fine-tuning on the preset neural network model according to the preprocessed device data to obtain a target device health detection model; Acquire current device data of the target track lighting device, and perform device health detection processing on the current device data through the target device health detection model to obtain the current device health value of the target track lighting device; A device processing strategy is matched according to the current device health value to obtain a target device processing strategy, and device processing is performed on the target track lighting device according to the target device processing strategy.

[0007] Optionally, in the detection method of track lighting equipment, the preprocessing includes a first preprocessing and a second preprocessing; the first preprocessing includes filtering processing, normalization processing and data cleaning processing; the second preprocessing includes data missing determination processing and data filling processing; the historical equipment data includes historical equipment parameters, historical environmental parameters and historical fault data; the preprocessed equipment data includes first preprocessed equipment data and second preprocessed equipment data; The acquiring of a plurality of historical device data of a plurality of track lighting devices and preprocessing the plurality of historical device data to obtain preprocessed device data specifically includes: Determine a plurality of track lighting devices in a target track carriage, and obtain the historical device parameters, the historical environmental parameters, and the historical fault data of the plurality of track lighting devices within a preset period; The historical device parameters include device voltage, device current, device cooling fan speed, device power operation time, and device operating frequency; The historical environmental parameters include the ambient temperature, ambient humidity and altitude of the device; Performing filtering, normalization, and data cleaning on the historical device parameters and the historical environment parameters to obtain the first pre-processed device data; Data missing determination processing and data filling processing are performed on the historical fault data to obtain the second pre-processing equipment data.

[0008] Optionally, the detection method of the track lighting equipment, wherein the determining of the preset neural network model and the model training and model fine-tuning of the preset neural network model according to the pre-processed device data to obtain the target device health detection model, further comprises: Performing feature extraction processing on the first preprocessing device data and the second preprocessing device data to obtain a plurality of feature vectors; Data division processing is performed on the multiple feature vectors to obtain training set data, verification set data and test set data.

[0009] Optionally, the detection method of the track lighting equipment, wherein the determining of a preset neural network model, and performing model training and model fine-tuning on the preset neural network model according to the pre-processed equipment data to obtain a target equipment health detection model, specifically includes: Determine a preset neural network model, and perform model training processing on the preset neural network model according to the training set data to obtain a first device health detection model; Fine-tune the first device health detection model according to the validation set data to obtain a second device health detection model; Performing parameter tuning processing on the second device health detection model according to the test set data to obtain a target device health detection model; The model parameters of the target device health detection model include device standard value status, device deviation status, device threshold, device standard deviation, device action times and device action duration.

[0010] Optionally, in the detection method of track lighting equipment, the preset neural network model includes an input layer, a convolution layer, a pooling layer, a fully connected layer and an output layer; The performing model training processing on the preset neural network model according to the training set data to obtain a first device health detection model specifically includes: The training set data is sequentially input into the input layer, the convolution layer and the pooling layer of the preset neural network model, and the training set data is subjected to convolution processing and pooling processing respectively by the convolution layer and the pooling layer to obtain a feature map; The feature map is sequentially input into the fully connected layer and the output layer, and the feature map is subjected to nonlinear transformation processing and activation processing respectively by the fully connected layer and the output layer to generate feature results, and the first device health detection model is obtained.

[0011] Optionally, the detection method of the track lighting device, wherein the step of acquiring current device data of the target track lighting device and performing device health detection processing on the current device data through the target device health detection model to obtain the current device health value of the target track lighting device, specifically includes: Acquire current device data of the target track lighting device, and perform the first preprocessing and the feature extraction processing on the current device data to obtain a current feature vector; The current feature vector is input into the target device health detection model, and the target device health detection model is used to perform device health detection processing on the current device data to obtain a current device health value.

[0012] Optionally, the detection method of the track lighting device, wherein the device processing strategy is matched according to the current device health value to obtain a target device processing strategy, and the device processing is performed on the target track lighting device according to the target device processing strategy, specifically comprising: Determine a preset processing strategy database, and obtain a plurality of device health status intervals in the preset processing strategy database and a device processing strategy corresponding to each of the device health status intervals; Matching the current device health value with a plurality of device health status intervals in the preset processing strategy database to obtain a target device health status interval corresponding to the current device health value in the plurality of device health status intervals; Acquire a target device processing strategy corresponding to the target device health status interval, and perform device processing on the target track lighting device according to the target device processing strategy; The equipment processing includes any one of no processing required, equipment maintenance or equipment replacement.

[0013] In addition, to achieve the above-mentioned purpose, the present invention further provides a detection system for track lighting equipment, wherein the detection system for track lighting equipment comprises: A historical device data preprocessing module, used for acquiring a plurality of historical device data of a plurality of track lighting devices, and preprocessing the plurality of the historical device data to obtain preprocessed device data; A health detection model building module is used to determine a preset neural network model, and perform model training and model fine-tuning on the preset neural network model according to the pre-processed device data to obtain a target device health detection model; An equipment health detection module is used to obtain current equipment data of a target track lighting device, and perform equipment health detection processing on the current equipment data through the target equipment health detection model to obtain a current equipment health value of the target track lighting device; The device processing strategy matching module is used to match the device processing strategy according to the current device health value, obtain the target device processing strategy, and perform device processing on the target track lighting device according to the target device processing strategy.

[0014] In addition, to achieve the above-mentioned purpose, the present invention also provides a terminal, wherein the terminal includes: a memory, a processor, and a detection program for track lighting equipment stored in the memory and executable on the processor, and when the detection program for track lighting equipment is executed by the processor, the steps of the detection method for track lighting equipment as described above are implemented.

[0015] In addition, to achieve the above-mentioned purpose, the present invention also provides a computer-readable storage medium, wherein the computer-readable storage medium stores a detection program for track lighting equipment, and when the detection program for track lighting equipment is executed by a processor, the steps of the detection method for track lighting equipment as described above are implemented.

[0016] In the present invention, multiple historical device data of multiple track lighting devices are obtained, and the multiple historical device data are preprocessed to obtain preprocessed device data; a preset neural network model is determined, and the preset neural network model is model trained and fine-tuned according to the preprocessed device data to obtain a target device health detection model; the current device data of the target track lighting device is obtained, and the current device data is subjected to device health detection processing through the target device health detection model to obtain the current device health value of the target track lighting device; the device processing strategy is matched according to the current device health value to obtain the target device processing strategy, and the target track lighting device is subjected to device processing according to the target device processing strategy. The present invention can accurately obtain the current device status of the track lighting device by constructing a target device health detection model and calculating the health value of the track lighting device according to the target device health detection model; further, the device processing strategy is automatically called according to the health value of the track lighting device to perform health management and maintenance, which can effectively improve the reliability and service life of the track lighting device and effectively ensure the safe and efficient operation of rail transit. BRIEF DESCRIPTION OF THE DRAWINGS

[0017] Figure 1 is a flow chart of a preferred embodiment of a detection method for track lighting equipment of the present invention; Figure 2 is a structural diagram of a preferred embodiment of a detection system for a track lighting device of the present invention; Figure 3 It is a structural diagram of a preferred embodiment of the terminal of the present invention. DETAILED DESCRIPTION

[0018] In order to make the purpose, technical solution and advantages of the present invention clearer and more specific, the present invention is further described in detail below with reference to the accompanying drawings and examples. It should be understood that the specific embodiments described herein are only used to explain the present invention and are not used to limit the present invention.

[0019] With the rapid development of urban rail transit, the safety, reliability and efficiency of rail transit lighting equipment are receiving more and more attention. Traditional lighting equipment maintenance methods mainly rely on manual regular inspections, which have problems such as high labor costs, low inspection frequency and incomplete data collection, making it difficult to accurately assess the health status of lighting equipment. At the same time, there is a lack of health evaluation standards and repair solutions for lighting equipment, and it is impossible to perform timely and effective maintenance on abnormal equipment, which may lead to the expansion of equipment failures, thereby affecting the normal operation of rail transit.

[0020] To solve the above problems, the present invention proposes a detection method for track lighting equipment, which establishes a health evaluation model by comprehensively collecting the operating parameters and fault data of the track lighting equipment, and uses the health evaluation model to monitor and evaluate the health status of the track lighting equipment in real time. Furthermore, a corresponding maintenance strategy is formulated based on the evaluation results, thereby improving the reliability and service life of the track lighting equipment and ensuring the safe and efficient operation of rail transit.

[0021] The detection method of the track lighting equipment according to the preferred embodiment of the present invention is as follows: Figure 1 As shown, the detection method of the track lighting equipment includes the following steps: Step S10, obtaining multiple historical device data of multiple track lighting devices, and preprocessing the multiple historical device data to obtain preprocessed device data. The preprocessing includes a first preprocessing and a second preprocessing; the first preprocessing includes filtering, normalization, and data cleaning; the second preprocessing includes data missing determination and data filling; the historical device data includes historical device parameters, historical environmental parameters, and historical fault data; the preprocessed device data includes first preprocessed device data and second preprocessed device data.

[0022] Each track lighting device has a controller inside, which is used to collect device parameters and environmental parameters. The track lighting device may include a power supply or LED lamp beads for car lighting, and may also include a carrier, a light guide plate, a monitoring camera, a PM2.5 sensor, a CO sensor, a formaldehyde sensor, a smoke sensor, a temperature and humidity sensor, a light intensity sensor, a sound sensor, an air pressure sensor, a human body sensor, a voice microphone, a vibration sensor and a control circuit, etc. The present invention is preferably introduced with a power supply or LED lamp beads for car lighting.

[0023] After determining the track lighting equipment, it is necessary to obtain the historical equipment parameters, historical environmental parameters and historical fault data corresponding to these track lighting equipment. However, these data may contain some useless data, and the data format is not uniform. Therefore, after obtaining the historical equipment parameters, historical environmental parameters and historical fault data, it is also necessary to preprocess the data to reduce the noise and interference in these data, thereby improving the quality and stability of the data, so that standard data can be formed in the future, thereby constructing a data set to train the model.

[0024] Specifically, multiple track lighting devices in the target track car are determined, and the historical device parameters, the historical environmental parameters and the historical fault data of the multiple track lighting devices within a preset period are obtained; wherein the historical device parameters include device voltage, device current, device cooling fan speed, device power operation time and device operating frequency; the historical environmental parameters include the ambient temperature, ambient humidity and altitude of the device.

[0025] The historical device parameters include not only device voltage, device current, device cooling fan speed, device power running time, and device operating frequency, but also parameters such as timestamp, brightness, switch status, and running time. The historical fault data may include various abnormal data such as no input voltage, no output voltage, over-high temperature, fan damage, communication failure, actual use time greater than the designed life, LED light source damage (not bright), etc.

[0026] The historical equipment parameters and the historical environmental parameters are subjected to filtering, normalization and data cleaning to obtain the first pre-processed equipment data; the historical fault data are subjected to data missing determination and data filling to obtain the second pre-processed equipment data.

[0027] Among them, the processing process of the historical fault data is as follows: 1. Perform data missing judgment on the historical fault data to determine the missing data in the historical fault data; 2. Perform data filling processing on the missing data to obtain data after data filling processing; 3. It can also be set to perform outlier detection processing on the data after data filling processing to obtain the second pre-processing equipment data of the track lighting equipment.

[0028] Step S20: determine a preset neural network model, and perform model training and model fine-tuning on the preset neural network model according to the preprocessed device data to obtain a target device health detection model.

[0029] The present invention uses historical equipment parameters, historical environmental parameters and historical fault data to construct a data set to implement model training of a preset neural network model, thereby obtaining a target equipment health detection model, and calculating the equipment health of the track lighting equipment through the target equipment health detection model.

[0030] Specifically, feature extraction processing is performed on the first preprocessing device data and the second preprocessing device data to obtain multiple feature vectors; data partitioning processing is performed on the multiple feature vectors to obtain training set data, verification set data and test set data.

[0031] After obtaining the preprocessed first preprocessing device data and the second preprocessing device data, it is necessary to perform feature extraction processing on the first preprocessing device data and the second preprocessing device data (split the first preprocessing device data and the second preprocessing device data into vector form) and annotate them to construct a data set for model training. Among them, the present invention divides the first preprocessing device data and the second preprocessing device data into training set data, verification set data and test set data. Among them, the division ratio of the training set data, the verification set data and the test set data can be 70% of the training set data, 15% of the verification set data, and 15% of the test set data. This division method can ensure the generalization ability of the model on different data sets.

[0032] Determine a preset neural network model, which includes an input layer, a convolution layer, a pooling layer, a fully connected layer and an output layer; input the training set data into the input layer, the convolution layer and the pooling layer of the preset neural network model in sequence, and perform convolution processing and pooling processing on the training set data through the convolution layer and the pooling layer, respectively, to obtain a feature map; input the feature map into the fully connected layer and the output layer in sequence, and perform nonlinear transformation processing and activation processing on the feature map through the fully connected layer and the output layer, respectively, to generate a feature result, and obtain the first equipment health detection model.

[0033] The preset neural network model in the present invention uses machine learning algorithms, such as support vector machines, decision trees, etc., to train various parameters of track lighting equipment and establish a target equipment health detection model. The input parameters of the model include the constructed training set data, validation set data and test set data, and the output of the model is the equipment health value of the track lighting equipment.

[0034] The preset neural network model used in the present invention includes an input layer, a convolution layer, a pooling layer, a fully connected layer and an output layer. By inputting the training set data into the preset neural network model, the corresponding equipment health value result is generated. The training process can be as follows: the training set data is input into the input layer of the preset neural network model, and the feature vector is convolved and pooled respectively through the convolution layer and the pooling layer to obtain the feature map after convolution and pooling. Furthermore, the feature map after convolution and pooling is vector expanded and nonlinearly transformed through the fully connected layer to obtain the output feature map of the fully connected layer. Finally, the output feature map of the fully connected layer is activated through the output layer to generate the output result of the corresponding feature map.

[0035] It can be understood that in the present invention, the training set data is first input into the preset neural network model, and then the output of each node is solved according to the relationship between each neuron. When the loss of the training result is outside the predetermined loss range, it is necessary to further calculate the error of the neuron and update the calculation weight, and then calculate the output of each neuron node again until the convergence condition is met. After the convergence condition is met, the weight is fixed and input as a determination parameter in actual detection.

[0036] The first device health detection model is fine-tuned according to the validation set data to obtain a second device health detection model; the second device health detection model is parameter-tuned according to the test set data to obtain a target device health detection model; wherein the model parameters of the target device health detection model include device standard value status, device deviation status, device threshold, device standard deviation, device action times and device action duration.

[0037] After obtaining the first device health detection model, it is necessary to iteratively train the first device health detection model. The trained first device health detection model can be cross-validated through the validation set data to obtain the second device health detection model. After that, the cross-validated second device health detection model is tested through the test set data to obtain the target device health detection model.

[0038] In the process of performing parameter tuning on the cross-validated second device health detection model through the test set data, the fitness index of the model parameter group can be calculated through the fitness function, and the model can be trained through the fitness index to determine its performance, such as accuracy, recall rate, etc.

[0039] Step S30: Acquire current device data of the target track lighting device, and perform device health detection processing on the current device data through the target device health detection model to obtain the current device health value of the target track lighting device.

[0040] After the target device health detection model is constructed, real-time health analysis can be performed on each track lighting device based on the target device health detection model. It is only necessary to collect the current device data of the track lighting device that needs to be analyzed and input the current device data into the target device health detection model. The target device health detection model will then perform device health detection processing and analysis on the current device data, thereby obtaining the current device health value corresponding to the target track lighting device.

[0041] Specifically, current device data of the target track lighting device is acquired, and the first preprocessing and the feature extraction processing are performed on the current device data to obtain a current feature vector; the current feature vector is input into the target device health detection model, and device health detection processing is performed on the current device data through the target device health detection model to obtain a current device health value.

[0042] For example, if the track lighting equipment that currently needs to be analyzed for health status is an LED lamp bead, it is necessary to obtain the device parameter and environmental parameter data corresponding to the LED lamp bead. After data preprocessing and feature value extraction, the data can be input into the target device health detection model. The target device health detection model will perform health analysis on the input data and then output the health value corresponding to the LED lamp bead. The current device status of the LED lamp bead can be understood through the health value.

[0043] Step S40: Match the device processing strategy according to the current device health value to obtain the target device processing strategy, and perform device processing on the target track lighting device according to the target device processing strategy.

[0044] According to the current device health value of each track lighting device, it is determined to which device health status interval the current device health value belongs. For example, the device health status interval can be divided into a first interval, a second interval and a third interval according to the severity of the device failure, and it is determined to which device health status interval the current device health value of the track lighting device is. For example, if the current device health value of the current track lighting device belongs to the second interval, the device processing strategy corresponding to the second interval is extracted from the database, and the current track lighting device is processed according to the device processing strategy.

[0045] Specifically, a preset processing strategy database is determined, and multiple device health status intervals and device processing strategies corresponding to each device health status interval in the preset processing strategy database are obtained; the current device health value is matched with the multiple device health status intervals in the preset processing strategy database to obtain a target device health status interval corresponding to the current device health value in the multiple device health status intervals; a target device processing strategy corresponding to the target device health status interval is obtained, and device processing is performed on the target track lighting device according to the target device processing strategy; wherein the device processing includes any one of no processing required, equipment maintenance, or equipment replacement.

[0046] The device health status interval in the present invention includes a first health status interval, a second health status interval and a third health status interval; the device health status interval includes a first device processing strategy corresponding to the first health status interval, a second device processing strategy corresponding to the second health status interval and a third device processing strategy corresponding to the third health status interval.

[0047] Generally, the first health status interval can be set to 80-100 points (indicating high health), the second health status interval can be set to 60-79 points (indicating average health); the third health status interval can be set to less than 60 points (indicating poor health). Furthermore, the first equipment processing strategy corresponding to the first health status interval can be set to no processing required; the second equipment processing strategy corresponding to the second health status interval can be set to equipment health maintenance; the third equipment processing strategy corresponding to the third health status interval can be set to equipment replacement.

[0048] For example, when the equipment health value is below 60 points, the equipment is determined to be in an irreparable state (i.e., in the third health state range), and the equipment needs to be replaced; when the equipment health value is between 60-79 points (i.e., in the second health state range), the equipment is determined to be repairable and requires equipment health inspection; when the equipment health value is 80 points or above (i.e., in the first health state range), the equipment is determined to be normal and no processing is required.

[0049] For example, when the life of the carriage lighting power supply in the track lighting equipment is about to end, it can be detected that the equipment health value of the carriage lighting power supply is in the third health state range, and an alarm prompt will be automatically generated. At the same time, a corresponding equipment replacement strategy will also be generated. The staff will check the alarm prompt and replace the carriage lighting power supply that is about to expire according to the equipment replacement strategy, effectively ensuring the subsequent normal operation of the carriage lighting power supply.

[0050] Furthermore, for track lighting equipment that is in the second health status interval and the third health status interval, health information data of these abnormal working states can be collected and compared with the working status health information stored in the fault database to determine whether the faults of the track lighting equipment in these abnormal working states belong to the fault categories stored in the fault database. If so, the processing plan corresponding to the fault category is output, and the processing results, fault causes and solutions are stored in the fault database to complete the update of the processing plan in the fault database; if it is not a stored fault category, the health information data of the abnormal working state of the track lighting equipment is saved, and the processing plan for notification inspection is output. After the inspection is completed, a new fault category is established in the fault database based on the inspection results and the processing plan, and the fault category is classified into the corresponding processing plan. At the same time, the cause of the fault is recorded to complete the update of the fault category in the fault database.

[0051] In addition, for track lighting equipment that is found to be abnormal, the cause of the abnormality can be further analyzed. If it is caused by external interference (such as interference from other equipment in the car, excessively high ambient temperature, excessively high humidity, excessively high altitude, and vibration factors, etc.), corresponding treatment measures can be taken according to these abnormal causes. For example, when the cause of the abnormality of the track lighting equipment is interference from other equipment in the car, the track lighting equipment can be electromagnetically shielded, or the installation position of the track lighting equipment can be adjusted; when the cause of the abnormality is excessively high ambient temperature, a heat dissipation device can be added or the working mode of the track lighting equipment can be adjusted; when the cause of the abnormality is excessively high humidity, moisture-proof measures can be taken; when the cause of the abnormality is excessively high altitude, the working voltage of the lighting equipment can be adjusted; when the cause of the abnormality is the influence of vibration, the shock-absorbing support of the lighting equipment can be strengthened.

[0052] Furthermore, the present invention will monitor the health status of the track lighting equipment in real time. When it is found that the fault information of the track lighting equipment that requires emergency treatment has returned to normal, it is necessary to send a fault recovery message to the lighting management system so that the lighting management system can update the status of the track lighting equipment and push it to the mobile terminal of the corresponding maintenance personnel for prompting.

[0053] The present invention uses a push module independent of the equipment status monitoring system to send alarm messages, realizes asynchronous control, ensures that the main process of the equipment status monitoring system for statistically analyzing fault information is not interfered with, and can also realize a timeout retransmission mechanism to avoid missing alarm messages.

[0054] In addition, the present invention can also use real-time stream analysis technology to determine the equipment health indicator parameter values ​​corresponding to each equipment status data, thereby improving the timeliness of data analysis. By providing multiple health standard parameter dimensions such as equipment standard value status, deviation status, threshold, standard deviation, equipment action times, and equipment action duration, comprehensive visualization and graphical management of these track lighting devices to be managed can be achieved.

[0055] Beneficial effects of the present invention: 1. It can realize real-time monitoring and evaluation of the health status of track lighting equipment, detect abnormal equipment in time and take corresponding maintenance measures, thus improving the reliability and service life of track lighting equipment; 2. By collecting historical equipment parameters, historical environmental parameters and historical fault data of lighting equipment, a scientific target equipment health detection model is established, which overcomes the defects of traditional reliance on manual regular inspections and improves the comprehensiveness of data collection and the accuracy of evaluation; 3. According to track lighting equipment with different health levels, corresponding repair and treatment plans have been formulated, including equipment health inspection or equipment replacement, which can provide timely and effective maintenance for abnormal equipment to avoid the expansion of equipment failures; 4. The health evaluation model of the present invention has high accuracy and reliability, and the evaluation results are more accurate; 5. The present invention adopts an intelligent health assessment method, which can achieve accurate assessment of the health status of lighting equipment and improve the efficiency and level of assessment.

[0056] Furthermore, if Figure 2 As shown, based on the above-mentioned detection method of track lighting equipment, the present invention also provides a detection system of track lighting equipment, wherein the detection system of track lighting equipment includes: A historical device data preprocessing module 51 is used to obtain a plurality of historical device data of a plurality of track lighting devices, and preprocess the plurality of the historical device data to obtain preprocessed device data; A health detection model building module 52 is used to determine a preset neural network model, and perform model training and model fine-tuning on the preset neural network model according to the pre-processed device data to obtain a target device health detection model; The device health detection module 53 is used to obtain the current device data of the target track lighting device, and perform device health detection processing on the current device data through the target device health detection model to obtain the current device health value of the target track lighting device; The device processing strategy matching module 54 is used to match the device processing strategy according to the current device health value, obtain the target device processing strategy, and perform device processing on the target track lighting device according to the target device processing strategy.

[0057] Furthermore, if Figure 3 As shown, based on the detection method and system of the track lighting equipment, the present invention also provides a terminal accordingly, and the terminal includes a processor 10, a memory 20 and a display 30. Figure 3 Only some components of the terminal are shown, but it should be understood that it is not required to implement all of the components shown, and more or fewer components may be implemented instead.

[0058] In some embodiments, the memory 20 may be an internal storage unit of the terminal, such as a hard disk or memory of the terminal. In other embodiments, the memory 20 may also be an external storage device of the terminal, such as a plug-in hard disk, a smart media card (SMC), a secure digital (SD) card, a flash card, etc. equipped on the terminal. Furthermore, the memory 20 may include both an internal storage unit of the terminal and an external storage device. The memory 20 is used to store application software and various types of data installed on the terminal, such as the program code of the installation terminal. The memory 20 may also be used to temporarily store data that has been output or is to be output. In one embodiment, a detection program 40 for a track lighting device is stored on the memory 20, and the detection program 40 for the track lighting device can be executed by the processor 10, thereby realizing the detection method for the track lighting device in the present application.

[0059] In some embodiments, the processor 10 may be a central processing unit (CPU), a microprocessor or other data processing chip, and is used to run the program code or process data stored in the memory 20, such as executing the detection method of the track lighting device.

[0060] In some embodiments, the display 30 may be an LED display, a liquid crystal display, a touch-sensitive liquid crystal display, an OLED (Organic Light-Emitting Diode) touch device, etc. The display 30 is used to display information on the terminal and to display a visual user interface.

[0061] In one embodiment, when the processor 10 executes the detection program 40 of the track lighting device in the memory 20, the following steps are implemented: Acquire multiple historical device data of multiple track lighting devices, and pre-process the multiple historical device data to obtain pre-processed device data; Determine a preset neural network model, and perform model training and model fine-tuning on the preset neural network model according to the preprocessed device data to obtain a target device health detection model; Acquire current device data of the target track lighting device, and perform device health detection processing on the current device data through the target device health detection model to obtain the current device health value of the target track lighting device; A device processing strategy is matched according to the current device health value to obtain a target device processing strategy, and device processing is performed on the target track lighting device according to the target device processing strategy.

[0062] Wherein, the preprocessing includes a first preprocessing and a second preprocessing; the first preprocessing includes filtering processing, normalization processing and data cleaning processing; the second preprocessing includes data missing determination processing and data filling processing; the historical equipment data includes historical equipment parameters, historical environmental parameters and historical fault data; the preprocessed equipment data includes first preprocessed equipment data and second preprocessed equipment data; The acquiring of a plurality of historical device data of a plurality of track lighting devices and preprocessing the plurality of historical device data to obtain preprocessed device data specifically includes: Determine a plurality of track lighting devices in a target track carriage, and obtain the historical device parameters, the historical environmental parameters, and the historical fault data of the plurality of track lighting devices within a preset period; The historical device parameters include device voltage, device current, device cooling fan speed, device power operation time, and device operating frequency; The historical environmental parameters include the ambient temperature, ambient humidity and altitude of the device; Performing filtering, normalization, and data cleaning on the historical device parameters and the historical environment parameters to obtain the first pre-processed device data; Data missing determination processing and data filling processing are performed on the historical fault data to obtain the second pre-processing equipment data.

[0063] The step of determining a preset neural network model and performing model training and model fine-tuning on the preset neural network model according to the pre-processed device data to obtain a target device health detection model may also include: Performing feature extraction processing on the first preprocessing device data and the second preprocessing device data to obtain a plurality of feature vectors; Data division processing is performed on the multiple feature vectors to obtain training set data, verification set data and test set data.

[0064] The step of determining a preset neural network model and performing model training and model fine-tuning on the preset neural network model according to the pre-processed device data to obtain a target device health detection model specifically includes: Determine a preset neural network model, and perform model training processing on the preset neural network model according to the training set data to obtain a first device health detection model; Fine-tune the first device health detection model according to the validation set data to obtain a second device health detection model; Performing parameter tuning processing on the second device health detection model according to the test set data to obtain a target device health detection model; The model parameters of the target device health detection model include device standard value status, device deviation status, device threshold, device standard deviation, device action times and device action duration.

[0065] The preset neural network model includes an input layer, a convolution layer, a pooling layer, a fully connected layer and an output layer; The performing model training processing on the preset neural network model according to the training set data to obtain a first device health detection model specifically includes: The training set data is sequentially input into the input layer, the convolution layer and the pooling layer of the preset neural network model, and the training set data is subjected to convolution processing and pooling processing respectively by the convolution layer and the pooling layer to obtain a feature map; The feature map is sequentially input into the fully connected layer and the output layer, and the feature map is subjected to nonlinear transformation processing and activation processing respectively by the fully connected layer and the output layer to generate feature results, and the first device health detection model is obtained.

[0066] The step of acquiring the current device data of the target track lighting device and performing device health detection processing on the current device data through the target device health detection model to obtain the current device health value of the target track lighting device specifically includes: Acquire current device data of the target track lighting device, and perform the first preprocessing and the feature extraction processing on the current device data to obtain a current feature vector; The current feature vector is input into the target device health detection model, and the target device health detection model is used to perform device health detection processing on the current device data to obtain a current device health value.

[0067] The matching of the device processing strategy according to the current device health value to obtain the target device processing strategy, and performing device processing on the target track lighting device according to the target device processing strategy, specifically includes: Determine a preset processing strategy database, and obtain a plurality of device health status intervals in the preset processing strategy database and a device processing strategy corresponding to each of the device health status intervals; Matching the current device health value with a plurality of device health status intervals in the preset processing strategy database to obtain a target device health status interval corresponding to the current device health value in the plurality of device health status intervals; Acquire a target device processing strategy corresponding to the target device health status interval, and perform device processing on the target track lighting device according to the target device processing strategy; The equipment processing includes any one of no processing required, equipment maintenance or equipment replacement.

[0068] The present invention also provides a computer-readable storage medium, wherein the computer-readable storage medium stores a detection program for a track lighting device, and when the detection program for the track lighting device is executed by a processor, the steps of the detection method for the track lighting device as described above are implemented.

[0069] In summary, the present invention provides a detection method and related equipment for track lighting equipment, the method comprising: obtaining multiple historical equipment data of multiple track lighting equipment, and preprocessing the multiple historical equipment data to obtain preprocessed equipment data; determining a preset neural network model, and performing model training and model fine-tuning on the preset neural network model according to the preprocessed equipment data to obtain a target equipment health detection model; obtaining the current equipment data of the target track lighting equipment, and performing equipment health detection processing on the current equipment data through the target equipment health detection model to obtain the current equipment health value of the target track lighting equipment; matching the equipment processing strategy according to the current equipment health value to obtain the target equipment processing strategy, and performing equipment processing on the target track lighting equipment according to the target equipment processing strategy. The present invention can accurately obtain the current equipment status of the track lighting equipment by constructing a target equipment health detection model and calculating the health value of the track lighting equipment according to the target equipment health detection model; further, automatically calling the equipment processing strategy according to the health value of the track lighting equipment to perform health management and maintenance, which can effectively improve the reliability and service life of the track lighting equipment, and effectively ensure the safe and efficient operation of rail transit.

[0070] It should be noted that, in this article, the terms "include", "comprises" or any other variations thereof are intended to cover non-exclusive inclusion, so that a process, method, article or terminal including a series of elements includes not only those elements, but also other elements not explicitly listed, or also includes elements inherent to such process, method, article or terminal. In the absence of further restrictions, an element defined by the sentence "comprises a ..." does not exclude the presence of other identical elements in the process, method, article or terminal including the element.

[0071] Of course, those skilled in the art can understand that all or part of the processes in the above-mentioned embodiments can be implemented by instructing related hardware (such as a processor, a controller, etc.) through a computer program, and the program can be stored in a computer-readable storage medium that can be read by a computer, and the program can include the processes of the above-mentioned method embodiments when executed. The computer-readable storage medium can be a memory, a disk, an optical disk, etc.

[0072] It should be understood that the application of the present invention is not limited to the above examples. For ordinary technicians in this field, improvements or changes can be made based on the above description. All these improvements and changes should fall within the scope of protection of the claims attached to the present invention.

Claims

1. A detection method for track lighting equipment, characterized in that: The detection method of the track lighting device comprises: Acquire multiple historical device data of multiple track lighting devices, and pre-process the multiple historical device data to obtain pre-processed device data; Determine a preset neural network model, and perform model training and model fine-tuning on the preset neural network model according to the preprocessed device data to obtain a target device health detection model; Acquire current device data of the target track lighting device, and perform device health detection processing on the current device data through the target device health detection model to obtain the current device health value of the target track lighting device; A device processing strategy is matched according to the current device health value to obtain a target device processing strategy, and device processing is performed on the target track lighting device according to the target device processing strategy.

2. The detection method of track lighting equipment according to claim 1, characterized in that: The preprocessing includes a first preprocessing and a second preprocessing; the first preprocessing includes filtering processing, normalization processing and data cleaning processing; The second preprocessing includes data missing determination processing and data filling processing; The historical equipment data includes historical equipment parameters, historical environmental parameters and historical fault data; the pre-processed equipment data includes first pre-processed equipment data and second pre-processed equipment data; The acquiring of a plurality of historical device data of a plurality of track lighting devices and preprocessing the plurality of historical device data to obtain preprocessed device data specifically includes: Determine a plurality of track lighting devices in a target track carriage, and obtain the historical device parameters, the historical environmental parameters, and the historical fault data of the plurality of track lighting devices within a preset period; The historical device parameters include device voltage, device current, device cooling fan speed, device power operation time, and device operating frequency; The historical environmental parameters include the ambient temperature, ambient humidity and altitude of the device; Performing filtering, normalization, and data cleaning on the historical device parameters and the historical environment parameters to obtain the first pre-processed device data; Data missing determination processing and data filling processing are performed on the historical fault data to obtain the second pre-processing equipment data.

3. The detection method of track lighting equipment according to claim 2, characterized in that: The step of determining a preset neural network model and performing model training and model fine-tuning on the preset neural network model according to the pre-processed device data to obtain a target device health detection model may also include: Performing feature extraction processing on the first preprocessing device data and the second preprocessing device data to obtain a plurality of feature vectors; Data division processing is performed on the multiple feature vectors to obtain training set data, verification set data and test set data.

4. The detection method of track lighting equipment according to claim 3, characterized in that: The determining of a preset neural network model, and performing model training and model fine-tuning on the preset neural network model according to the pre-processed device data to obtain a target device health detection model specifically includes: Determine a preset neural network model, and perform model training processing on the preset neural network model according to the training set data to obtain a first device health detection model; Fine-tune the first device health detection model according to the validation set data to obtain a second device health detection model; Performing parameter tuning processing on the second device health detection model according to the test set data to obtain a target device health detection model; The model parameters of the target device health detection model include device standard value status, device deviation status, device threshold, device standard deviation, device action times and device action duration.

5. The detection method of track lighting equipment according to claim 4, characterized in that: The preset neural network model includes an input layer, a convolutional layer, a pooling layer, a fully connected layer and an output layer; The performing model training processing on the preset neural network model according to the training set data to obtain a first device health detection model specifically includes: The training set data is sequentially input into the input layer, the convolution layer and the pooling layer of the preset neural network model, and the training set data is convolutionally processed and pooled by the convolution layer and the pooling layer, respectively, to obtain a feature map; The feature map is sequentially input into the fully connected layer and the output layer, and the feature map is subjected to nonlinear transformation processing and activation processing respectively by the fully connected layer and the output layer to generate feature results, and the first device health detection model is obtained.

6. The detection method of track lighting equipment according to claim 3, characterized in that: The acquiring current device data of the target track lighting device, and performing device health detection processing on the current device data through the target device health detection model to obtain the current device health value of the target track lighting device specifically includes: Acquire current device data of the target track lighting device, and perform the first preprocessing and the feature extraction processing on the current device data to obtain a current feature vector; The current feature vector is input into the target device health detection model, and the target device health detection model is used to perform device health detection processing on the current device data to obtain a current device health value.

7. The detection method of track lighting equipment according to claim 1, characterized in that: The matching of the device processing strategy according to the current device health value to obtain the target device processing strategy, and performing device processing on the target track lighting device according to the target device processing strategy, specifically includes: Determine a preset processing strategy database, and obtain a plurality of device health status intervals in the preset processing strategy database and a device processing strategy corresponding to each of the device health status intervals; Matching the current device health value with a plurality of device health status intervals in the preset processing strategy database to obtain a target device health status interval corresponding to the current device health value in the plurality of device health status intervals; Acquire a target device processing strategy corresponding to the target device health status interval, and perform device processing on the target track lighting device according to the target device processing strategy; The equipment processing includes any one of no processing required, equipment maintenance or equipment replacement.

8. A detection system for track lighting equipment, characterized in that: The detection system of the track lighting equipment comprises: A historical device data preprocessing module, used for acquiring a plurality of historical device data of a plurality of track lighting devices, and preprocessing the plurality of the historical device data to obtain preprocessed device data; A health detection model building module is used to determine a preset neural network model, and perform model training and model fine-tuning on the preset neural network model according to the pre-processed device data to obtain a target device health detection model; An equipment health detection module is used to obtain current equipment data of a target track lighting device, and perform equipment health detection processing on the current equipment data through the target equipment health detection model to obtain a current equipment health value of the target track lighting device; The device processing strategy matching module is used to match the device processing strategy according to the current device health value, obtain the target device processing strategy, and perform device processing on the target track lighting device according to the target device processing strategy.

9. A terminal, characterized in that: The terminal includes: a memory, a processor, and a detection program for track lighting equipment stored in the memory and executable on the processor. When the detection program for track lighting equipment is executed by the processor, the steps of the detection method for track lighting equipment as described in any one of claims 1 to 7 are implemented.

10. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores a detection program for a track lighting device, and when the detection program for the track lighting device is executed by a processor, the steps of the detection method for the track lighting device according to any one of claims 1 to 7 are implemented.

Images