Fault prediction method and system for self-service luggage check-in equipment, medium and product

Abstract

The invention discloses a self-service luggage check-in equipment fault prediction method and system, a medium and a product. According to the method, a time sequence data sequence is constructed and divided by obtaining an operation data fragment of an internal sensor of equipment and an environment data fragment of an external sensor, the time sequence data sequence is input into an interference detection model, interfered data is corrected, then a correction time data sequence is constructed, and then a fault prediction model is input to obtain a result. And if the equipment is predicted to have a fault, positioning a fault risk part and sending a maintenance request to maintenance personnel. According to the scheme, through comprehensive data acquisition, data integrity guarantee and accurate fault prediction, the fault prediction accuracy and maintenance efficiency are effectively improved, the influence of external interference on a fault prediction result is reduced, and accurate equipment maintenance and fault prevention are realized.

Description

Technical Field

[0001] The present application relates to the field of intelligent device monitoring and maintenance technology, and in particular to a method, system, medium, and product for predicting faults of self-service baggage check-in equipment. Background Art

[0002] With the rapid development of the air transportation industry, self-service baggage drop-off systems have become an indispensable feature of modern airports. These systems provide efficient and convenient baggage check-in services, significantly reducing queues and the need for manual operation. To ensure the efficient operation of self-service baggage drop-off systems, airports typically perform regular maintenance to reduce equipment failure rates. However, with the continued growth in passenger volume during peak holiday periods, the operational pressure on self-service baggage drop-off systems has increased significantly, placing higher demands on their reliability and maintenance management.

[0003] To meet these requirements, related technologies typically monitor the operating status of equipment to identify potential faults. For example, some solutions collect operational data from self-service baggage check-in devices, including baggage weighing data, conveyor belt operating status data, and code scanning data, and use preset thresholds to determine whether the equipment is abnormal. Once the data is detected to be outside the threshold range, an alarm signal is triggered, alerting maintenance personnel to conduct inspections and maintenance. In addition, there are also technical solutions that use regular manual inspections to check the appearance and operating status of the equipment to avoid service interruptions caused by equipment failures.

[0004] However, these technologies can face challenges in predicting faults in the complex environments of self-service baggage drop equipment during peak hours at airports. External interference (such as collisions or stacking during baggage transport) with the equipment's operational and environmental data can lead to data collection anomalies, reducing the accuracy of fault predictions. In these situations, there's room for improvement in these technologies to provide maintenance personnel with timely and accurate guidance. Summary of the Invention

[0005] The present application provides a self-service baggage check-in equipment fault prediction method, system, medium and product, which are used to solve the problem of data being subject to external interference during the fault prediction process of the self-service baggage check-in equipment, thereby reducing the impact of external interference on the fault prediction results of the self-service baggage check-in equipment.

[0006] In a first aspect, a method for predicting a fault of a self-service baggage check-in device includes: obtaining an operation data segment generated during a baggage check-in process by an internal sensor of the self-service baggage check-in device, the operation data segment being a data segment of a preset internal sensor acquisition length; obtaining an environmental data segment of the surrounding area collected by an external sensor of the self-service baggage check-in device, the environmental data segment being a data segment of a preset external sensor acquisition length; constructing the operation data segment and the environmental data segment into a time series data sequence; dividing the time series data sequence into a plurality of time series data segments of a preset length; inputting the time series data segment into an interference detection model to determine whether the time series data segment is interfered with; if interfered with, performing data correction on the time series data segment to obtain a corrected data segment; constructing a corrected time data sequence based on the undisturbed time series data segment and the corrected data segment; inputting the corrected time data sequence into a fault prediction model to obtain a fault prediction result of the self-service baggage check-in device; and if the fault prediction result indicates that the self-service baggage check-in device will fail, sending a maintenance request for the fault risk location to maintenance personnel, the fault risk location being the fault location corresponding to the predicted failure of the self-service baggage check-in device.

[0007] By adopting the above technical solution, first, internal sensors collect operational data segments and external sensors collect environmental data segments, enabling the system to comprehensively capture dynamic information about equipment operation and the external environment. Second, this data is constructed into a time series data sequence and divided into time series data segments of preset length, ensuring that the time series characteristics are effectively preserved and enabling the model to capture time-dependent characteristics. Third, the interference detection model determines whether the data segments are interfered with and corrects the interfered data. This process ensures the integrity and accuracy of the data. Finally, the corrected time data sequence is input into the fault prediction model, which can more accurately predict the equipment's fault status and locate the fault risk area. In combination with maintenance requests, maintenance personnel are promptly notified, thereby achieving accurate maintenance and fault prevention of the equipment. This method significantly improves the accuracy of fault prediction and the efficiency of maintenance, avoiding misjudgments caused by external interference with the data.

[0008] In combination with some embodiments of the first aspect, in some embodiments, the training process of the interference detection model is: performing feature extraction on sample time series data segments to obtain operation data features and environmental data features, the operation data features include luggage weighing data, barcode scanning data and motion status data, and the environmental data features include temperature data, humidity data and air pressure data; constructing a time series data segment training data set containing multiple scenarios and multiple working conditions, the time series data segment training data set includes sample time series data segments and interference label annotations, the interference label includes an "interfered" label and an "undisturbed" label; training the interference detection model based on the time series data segment training data set.

[0009] By adopting the above technical solution, first, by extracting operational data features and environmental data features from sample time series data segments, the model can simultaneously focus on the correlation between the equipment's operating status and the external environment. Second, by constructing a training dataset of time series data segments covering multiple scenarios and multiple working conditions, and labeling the data as "interfered" or "undisturbed," the model can learn to identify interference features under different working conditions. Finally, based on the training data, the interference detection model is trained, enabling the model to extract interference features from the input time series data and make accurate judgments. This training method ensures the robustness and generalization ability of the interference detection model in complex scenarios, enabling it to accurately identify interference data and lay a solid foundation for subsequent data correction and fault prediction.

[0010] In combination with some embodiments of the first aspect, in some embodiments, the training process of the fault prediction model is: obtaining historical operation data, historical environmental data and historical fault records collected by the self-service baggage check-in device during operation, the historical operation data includes baggage weighing data, barcode scanning data and motion status data, the historical environmental data includes temperature data, humidity data and air pressure data, and the historical fault records include operation data and environmental data when the fault occurred; constructing a historical data training data set based on the historical operation data, the historical environmental data and the historical fault records; and training the fault prediction model based on the historical data training data set.

[0011] By adopting the above technical solution, first, by acquiring historical operating data, environmental data, and fault records collected during equipment operation and constructing this data into a training dataset, the training data fully covers the characteristics of the equipment under normal and abnormal conditions. Secondly, through the training of the fault prediction model, the model can learn the complex relationship between equipment operation and environmental conditions, as well as the impact of these conditions on the occurrence of faults. Finally, after sufficient training, the fault prediction model can more accurately determine whether the equipment has potential fault hazards. In this way, the fault prediction model not only improves prediction accuracy, but also effectively reduces the possibility of false positives and false negatives, thereby providing reliable technical support for the safe operation of the equipment.

[0012] In combination with some embodiments of the first aspect, in some embodiments, if the time series data segment is interfered with, data correction is performed on the time series data segment to obtain a corrected data segment, specifically including: if the time series data segment is interfered with, the time series data segment is compared with a preset normal data range to obtain an abnormal value that exceeds the normal data range; the time step of the abnormal value in the time series data segment is marked; if the abnormal value is a missing value caused by data acquisition loss, based on the normal data before and after the abnormal value in the time series data segment, a corrected value of the abnormal value is calculated by a linear interpolation formula; and the corrected data segment of the time series data segment is obtained according to the correction value.

[0013] By adopting the above technical solution, the specific time step of the outlier can be quickly located by comparing the disturbed time series data fragment with the normal data range. Then, by marking the outliers, the target range of data correction is clarified. Finally, considering the data loss, the linear interpolation formula is used to calculate the corrected values of the missing values, thereby filling the data gaps. Finally, after all outliers are corrected, the corrected data fragments are generated, and the continuity and integrity of the time series data are guaranteed. Through these steps, the system can effectively compensate for the damage caused by the interference data to the time series characteristics, providing high-quality data input for subsequent fault prediction, and significantly improving the reliability and accuracy of the prediction.

[0014] In conjunction with some embodiments of the first aspect, in some embodiments, after marking the time step of the abnormal value in the time series data segment, the method further includes: if the time step where the abnormal value is located deviates from the normal range by a high amplitude for a short period of time, determining it as a short-term spike abnormality; and calculating the correction value for the spike abnormality using a sliding window average method, the calculation formula of the sliding window average method is: Should is the time before and after the time step t The data value of the time step, w is the size of the sliding window, and x t ' is the correction value of the abnormal value; By adopting the above technical solution, first, by determining abnormal situations with short-term, high-amplitude deviations from the normal range, short-term spike anomalies can be quickly identified. Second, the sliding window average method is used to calculate the correction value, so that the abnormal data can be smoothly replaced by the surrounding normal data, avoiding the interference of spike anomalies on the overall trend of the time series data. Finally, the corrected data segments can better preserve the stability and continuity of the time series, while eliminating the negative impact of short-term anomalies on the fault prediction model. Through this correction method, the system can effectively reduce the impact of spike anomalies on data quality, thereby improving the accuracy and stability of equipment fault prediction.

[0015] In combination with some embodiments of the first aspect, in some embodiments, the operation data fragments and the environmental data fragments are constructed into a time series data sequence, specifically including: extracting the receiving timestamp and acquisition timestamp of each received operation data and environmental data, the receiving timestamp is used to mark the receiving time of the operation data and the environmental data, and the acquisition timestamp is used to mark the acquisition time of the operation data and the environmental data; storing the received operation data and the environmental data in the data buffer of the corresponding sensor; subtracting the data of the receiving timestamp from the data of the acquisition timestamp to obtain a data delay value; if the data delay value exceeds a preset threshold, subtracting the data of the receiving timestamp from the data delay value to obtain a correction timestamp; according to the correction timestamp, sorting the operation data and the environmental data according to the correction timestamp to obtain a time series data sequence with unified timing.

[0016] By adopting the above technical solution, first, by extracting the reception and acquisition timestamps of operating data and environmental data and calculating the data delay value, the delay in data acquisition and transmission can be accurately measured. Then, for data with delay values exceeding the preset threshold, timestamp correction is performed to ensure the time alignment of the data. Finally, the data is sorted according to the corrected timestamps to construct a time series data sequence with unified timing, thereby eliminating the impact of data asynchrony or misordering on time series feature analysis. Through the above steps, the system can construct a time series data sequence with temporal consistency and accuracy, providing a high-quality data foundation for subsequent interference detection and fault prediction.

[0017] In combination with some embodiments of the first aspect, in some embodiments, if the fault prediction result is that the self-service baggage check-in device will fail, a maintenance request for the fault risk portion is sent to the maintenance personnel, specifically including: determining the range of the operating data and the environmental data of the self-service baggage check-in device under normal operating conditions by analyzing the historical operating data and the environmental data; if the fault prediction result is that the self-service baggage check-in device will fail, extracting data segments in the corrected time data sequence that exceed the range of the operating data and the environmental data under normal operating conditions to obtain abnormal data segments; determining the target sensor corresponding to the abnormal data segment based on the acquisition timestamp of the abnormal data segment; and determining the equipment component corresponding to the target sensor as the fault risk portion.

[0018] By adopting the above technical solution, first, by combining historical operating data and environmental data to analyze the normal operating range of the equipment, a benchmark is provided for fault risk identification. Second, by comparing the corrected time data series with the normal range, abnormal data segments that exceed the range can be quickly identified. Third, the corresponding sensor is located using the acquisition timestamp of the abnormal data segment, and the sensor is associated with a specific equipment component to determine the fault risk location. In this way, the system can accurately locate the specific component that may fail, significantly improving the accuracy of fault location and the targeted maintenance, and reducing the maintenance cost and repair time caused by unclear fault scope.

[0019] In a second aspect, an embodiment of the present application provides a self-service baggage check-in equipment failure prediction system, which includes: one or more processors and a memory; the memory is coupled to the one or more processors, the memory is used to store computer program code, the computer program code includes computer instructions, and the one or more processors call the computer instructions to enable the self-service baggage check-in equipment failure prediction system to perform the method described in the first aspect and any possible implementation method of the first aspect.

[0020] In a third aspect, an embodiment of the present application provides a computer-readable storage medium comprising instructions. When the instructions are executed on a self-service baggage check-in equipment failure prediction system, the self-service baggage check-in equipment failure prediction system executes the method described in the first aspect and any possible implementation of the first aspect.

[0021] In a fourth aspect, an embodiment of the present application provides a computer program product. When the computer program product is run on a self-service baggage check-in equipment failure prediction system, the self-service baggage check-in equipment failure prediction system performs the method described in the first aspect and any possible implementation method of the first aspect.

[0022] One or more technical solutions provided in the embodiments of this application have at least the following technical effects or advantages: 1. By acquiring operational data segments and environmental data segments and constructing a time series data sequence, the system can simultaneously obtain information on the dynamic changes in the equipment's operating status and the external environment through internal sensors collecting operational data segments and external sensors collecting environmental data segments. Secondly, constructing operational data and environmental data into a time series data sequence and dividing them into time series data segments of preset lengths ensures that the time dimension characteristics of the data are preserved and lays the foundation for subsequent interference detection and fault prediction. Finally, the uniformity and continuity of the time series data sequence enable the model to accurately capture time-related characteristics and trend changes, avoiding prediction bias caused by data fragmentation or loss. Through this process, the system effectively solves the problem of fault prediction errors caused by incomplete data or lack of time series characteristics, thereby achieving high-precision equipment fault prediction.

[0023] 2. Due to the use of data interference detection and data correction methods, first, the time series data segments are detected in conjunction with the interference detection model, which can accurately identify whether the data segments are subject to external interference and mark the interfered data. Secondly, the interfered data segments are targeted through the data correction method, specifically marking outliers, using linear interpolation combined with missing data, and using the sliding window average method to correct short-term spike anomalies, ensuring the continuity and integrity of the data segments. Finally, the corrected time series data segments and the undisturbed segments are reconstructed into a corrected time data sequence, thereby improving data quality and model input accuracy. Through this process, the system effectively solves the problem of fault prediction errors caused by data interference and outliers, and improves the accuracy and robustness of equipment fault prediction under complex working conditions.

[0024] 3. Due to the use of fault prediction and risk location methods, the fault prediction model analyzes the corrected time data series and can accurately determine whether the equipment has potential fault hazards. By combining historical operating data with the normal range of environmental data for threshold comparison, it can quickly identify abnormal data fragments that exceed the normal range. The target sensor is located by the timestamp of the abnormal data fragment and associated with the specific equipment component, accurately marking the fault risk location. Finally, by sending a maintenance request to the maintenance personnel, it ensures that the fault location can be maintained and repaired in a timely manner. Through this process, the system effectively solves the problem of the inability to accurately locate the fault location in related technologies, realizes efficient and targeted equipment maintenance, and significantly reduces equipment downtime and maintenance costs. BRIEF DESCRIPTION OF THE DRAWINGS

[0025] Figure 1 This is a schematic diagram of an application scenario of the self-service baggage check-in equipment failure prediction method in an embodiment of the present application; Figure 2This is a flow chart of a method for predicting failures of self-service baggage check-in equipment according to an embodiment of the present application; Figure 3 This is another flowchart of the method for predicting failures of self-service baggage check-in equipment according to an embodiment of the present application; Figure 4 This is a structural diagram of a self-service baggage check-in equipment fault prediction system provided in an embodiment of the present application. DETAILED DESCRIPTION

[0026] The terms used in the following examples of the present application are only for the purpose of describing specific embodiments and are not intended to limit the present application. As used in the specification and appended claims of this application, the singular expressions "a," "an," "said," "above," "the," and "this" are intended to include plural expressions as well, unless the context clearly indicates otherwise. It should also be understood that the term "and / or" used in this application refers to any or all possible combinations comprising one or more of the listed items.

[0027] In the following, the terms "first" and "second" are used for descriptive purposes only and should not be understood to imply or suggest relative importance or implicitly indicate the number of the technical features indicated. Therefore, the features defined as "first" and "second" may explicitly or implicitly include one or more of the features. In the description of the embodiments of this application, unless otherwise specified, "plurality" means two or more.

[0028] Before introducing the embodiments of the present application, some terms involved in the embodiments of the present application are first defined and explained.

[0029] Motion data: refers to the real-time status data collected by sensors during the operation of the device, such as displacement, velocity, acceleration, etc., which is used to reflect the dynamic behavior of the device.

[0030] Environmental data: refers to the external condition data collected from the device's environment, such as temperature, humidity, and air pressure, which is used to assess the impact of the external environment on device operation.

[0031] Time data series: refers to a set of continuous data arranged in sequence with time as the key index, which is used to represent the dynamic change trend and time correlation of data.

[0032] Figure 1 This is a schematic diagram of an application scenario of the self-service baggage check-in equipment failure prediction method in an embodiment of the present application.

[0033] See also Figure 1 The self-service baggage check-in device 100 includes internal sensors and external sensors.

[0034] 101. Internal sensors mainly include weight sensors, position sensors, speed sensors, vibration sensors, and barcode scanners. The data collected by internal sensors is operational data.

[0035] 102. External sensors mainly include temperature sensors, humidity sensors, air pressure sensors, light sensors, and noise sensors. The data collected by external sensors is environmental data.

[0036] Current methods for predicting failures in self-service baggage drop-off devices primarily analyze data collected by sensors within the devices and use the results to determine if the devices are faulty. However, data collected by internal and external sensors can be subject to external interference, which can affect the device's failure prediction results.

[0037] The present application embodiment provides a self-service baggage check-in equipment fault prediction method, which can reduce external interference with data in this case and thus improve the accuracy of fault prediction. The self-service baggage check-in equipment fault prediction method in the present application embodiment is described as follows: Figure 2 This is a flow chart of the method for predicting failures of self-service baggage check-in equipment using the embodiment of the present application. Figure 2 The self-service baggage check-in equipment fault prediction method is specifically described as follows: 201. Obtain an operation data segment generated during the baggage check-in process and collected by an internal sensor of the self-service baggage check-in equipment. The operation data segment is a data segment with a preset internal sensor collection length.

[0038] Obtain environmental data segments around the self-service baggage check-in device collected by external sensors, where the environmental data segments are data segments with preset external sensor collection lengths.

[0039] The self-service baggage drop device failure prediction system (hereinafter referred to as the system) first needs to obtain comprehensive and accurate data as the basis for subsequent analysis and prediction. The system collects relevant data from internal and external sensors of the self-service baggage drop device. For operational data captured by internal sensors, the system establishes stable data transmission channels with various internal sensors, such as weight sensors, position sensors, speed sensors, and motor current sensors. During the baggage check-in process, these sensors each perform their respective functions, continuously capturing critical information about the equipment's operation. For example, the weight sensor monitors the weight of the baggage in real time. From the moment the baggage is placed on the check-in scale, it updates data every millisecond, accurate to grams or even milligrams, providing the system with a dynamic picture of the baggage's weight. The speed sensor monitors the conveyor belt's speed, ensuring that the baggage is conveyed at an appropriate speed, neither too fast to cause the baggage to slip or overload the equipment, nor too slow to affect check-in efficiency. The data collected by these internal sensors is not transmitted in a random manner. Instead, it is captured by the system in data segments according to the preset internal sensor collection time. Assuming the preset collection time window is 5 seconds, the system will integrate the data generated by each internal sensor during this period every 5 seconds to form an operating data segment containing various operating information such as weight and speed, recording the dynamic operating status of the equipment during that specific time. The system also collects fragments of environmental data from external sensors. These sensors, including temperature, humidity, and air pressure sensors, are installed at strategic locations around the self-service bag drop machines to capture real-time environmental conditions. Temperature sensors constantly monitor ambient temperature. In the summer, airport terminal halls can be hot due to the dense crowds and heavy air conditioning loads, while in the winter, they can be cooler due to the infiltration of cool air from outside. These temperature fluctuations are precisely measured and transmitted to the system. Humidity sensors monitor ambient humidity. Airport environments can fluctuate widely depending on location, season, and ventilation conditions. High humidity can cause short circuits in the device's internal electronic components, while low humidity can cause static electricity, impacting normal operation. Pressure sensors record ambient air pressure. For example, at high-altitude airports, air pressure can differ significantly from that at lower altitudes, potentially impacting the performance of certain components, such as pressure-sensitive elements. Similarly, the data collected by these external sensors is also captured by the system in the form of data segments according to the preset external sensor acquisition time window. If the preset acquisition time window is also 5 seconds, then every 5 seconds the system will capture an environmental data segment containing environmental information such as temperature, humidity, and air pressure. In this way, the system has a comprehensive understanding of the device's own operating status data and the surrounding environmental impact data during operation. 202. Construct the operating data segments and the environmental data segments into a time series data sequence.

[0040] According to the time series data sequence, a plurality of time series data segments of preset lengths are obtained by division.

[0041] After obtaining the operational data fragments and environmental data fragments, the next key task of the self-service baggage check-in equipment failure prediction system is to effectively organize and process these data for subsequent analysis and model input. The system first constructs the operational data segments and environmental data segments into a time series data sequence. A time series data sequence is constructed by aligning the operational data segments with the environmental data segments according to their timestamps and combining them to form multidimensional time series data.

[0042] After constructing the time series data sequence, the system needs to further divide it to meet the needs of subsequent model processing. The system will divide the time series data sequence into several time series data segments of preset lengths based on the preset length.

[0043] 203. Input the time series data segment into an interference detection model to determine whether the time series data segment is interfered with.

[0044] After completing the data segmentation, the self-service baggage check-in equipment fault prediction system needs to ensure that the data input into the subsequent fault prediction model is reliable. Therefore, the time series data segments obtained by segmentation will be input into the interference detection model to determine whether these data segments are interfered with. The interference detection model is a crucial component of the entire fault prediction process. Built on a rule-based algorithm, it accurately identifies abnormal interference within time series data segments. Before being put into use, the model underwent a rigorous training process. The training process for the interference detection model is as follows: Extract features from the sample time series data segments to obtain operational data features and environmental data features. The operational data features include baggage weighing data, barcode scanning data, and motion status data. The environmental data features include temperature data, humidity data, and air pressure data. A training dataset of time series data segments from multiple scenarios and operating conditions is constructed. The training dataset includes sample time series data segments and interference labels, including "interfered" and "undisturbed" labels. The interference detection model is then trained based on this training dataset. During training, the model continuously learns the correlation between the characteristic patterns in the dataset and the interference labels. Through repeated iterative training, the model gradually becomes able to accurately determine whether newly input time series data segments are subject to interference. When the resulting time series data segments of a preset length are input into the interference detection model, the model rapidly analyzes them, extracting operational and environmental data features, and compares them with the normal and interference patterns learned during training. If the model determines that the features in the time series data segment have a high degree of match with the "interfered" pattern, the data segment is deemed to be subject to interference; otherwise, it is deemed to be undisturbed. Only time series data segments verified as undisturbed by the interference detection model are further used by the system for subsequent self-service baggage check-in equipment fault prediction, effectively reducing external data interference and improving fault prediction accuracy.

[0045] 204. If interference occurs, perform data correction on the time series data segment to obtain a corrected data segment.

[0046] When the self-service baggage check-in device fault prediction system determines, through its interference detection model, that a time series data segment has been disturbed, it immediately initiates a data correction process to ensure the quality of the data input into subsequent analysis. Based on the specific characteristics and actual circumstances of the disturbed data, the system determines the type of data interference and flexibly uses interpolation algorithms, prediction models, or historical data supplementation to correct the outliers in the disturbed data, thereby generating a corrected data segment.

[0047] 205. Construct a corrected time data sequence according to the undisturbed time series data segments and the corrected data segments.

[0048] After correcting the disturbed time series data segments, the self-service baggage drop device failure prediction system recombines the uninterrupted time series data segments with the corrected data segments to construct a corrected time series data sequence. This process is like piecing together a scrambled puzzle to form uninterrupted, continuous data, providing strong support for subsequent accurate analysis and prediction of self-service baggage drop device failures.

[0049] The system first categorizes and organizes all time series data segments that have undergone interference detection, placing those identified as undisturbed and those generated after data correction in separate temporary storage areas. The system then sorts these segments in chronological order based on their timestamp information. In actual operation, more complex situations may arise, such as when there are multiple disturbed data segments scattered across the time series. In these cases, the system utilizes efficient algorithms and data structures to quickly and accurately integrate all undisturbed and corrected segments.

[0050] By constructing a corrected time data series, the system eliminates interference data that disrupts the continuity and integrity of the time series, allowing subsequent analysis to be based on a stable and reliable data set. This continuous data series more realistically reflects the actual operating status of the self-service baggage drop equipment and changes in the surrounding environment over time, providing high-quality input data for the fault prediction model, helping to improve the accuracy and reliability of fault predictions.

[0051] 206. Input the corrected time data sequence into a fault prediction model to obtain a fault prediction result of the self-service baggage check-in device.

[0052] After a series of data collection, processing, and correction processes, the self-service baggage drop-off device failure prediction system ultimately inputs the constructed corrected time data series into a fault prediction model to determine whether the self-service baggage drop-off device is at risk of imminent failure. The fault prediction model, the core component of the entire system, is built using deep learning technology. It conducts in-depth analysis of the input corrected time data series, exploring the potential relationships between device operating status and environmental conditions, thereby accurately determining the risk of device failure. Before being put into use, the fault prediction model underwent an extensive training process. This training process involved acquiring historical operating data, historical environmental data, and historical fault records collected during the operation of the self-service baggage check-in device. The historical operating data included baggage weighing data, barcode scanning data, and motion status data; the historical environmental data included temperature data, humidity data, and air pressure data; and the historical fault records included the operating data and environmental data at the time the fault occurred.

[0053] The fault data in the historical operation data and the historical environment data are annotated according to the historical fault records, and a historical data training data set is constructed according to the annotated results.

[0054] This historical data training dataset is used to train the fault prediction model. During training, the model continuously adjusts its parameters to learn the characteristic patterns and fault occurrence patterns in the data. For example, through a large number of training samples, the model gradually grasps that in high temperature and high humidity environments, when baggage weighing data fluctuates abnormally and the barcode scanning error rate increases, the probability of equipment failure increases significantly. After long periods of training and multiple rounds, the model achieves a high level of prediction accuracy. When the corrected time data series is input into the trained fault prediction model, the model quickly analyzes the data and, based on the learned patterns, determines whether the self-service baggage drop equipment is at risk of imminent failure. The model then outputs a prediction result, providing airport operations and maintenance personnel with timely and accurate equipment failure warnings, allowing them to take proactive maintenance measures to ensure the stable operation of the self-service baggage drop equipment.

[0055] 207. If the fault prediction result indicates that the self-service baggage check-in device will fail, a maintenance request for the fault risk location is sent to a maintenance personnel. The fault risk location is the fault location corresponding to the predicted failure of the self-service baggage check-in device.

[0056] When the self-service baggage check-in equipment fault prediction system obtains a fault prediction result obtained through the fault prediction model and shows that the self-service baggage check-in equipment will fail, the system immediately enters the fault response process and sends a maintenance request for the fault-risk area to the maintenance personnel. This maintenance request contains extremely critical information, providing strong support for the maintenance personnel to carry out maintenance work efficiently.

[0057] 208. If the fault prediction result is that the self-service baggage check-in device will not fail, no response is made.

[0058] When the fault prediction results obtained by the self-service baggage check-in equipment fault prediction system through the fault prediction model show that the self-service baggage check-in equipment will not fail, on the surface, the system seems to require no additional operations. However, in fact, the system still plays an important monitoring and recording role in this process, laying the foundation for ensuring the long-term stable operation of the equipment. Although the device is currently predicted to be fault-free, the system does not relax its continuous monitoring of the device's operating status. It continues to collect real-time operational data from the self-service baggage drop device's internal sensors and environmental data from external sensors according to the established data collection frequency and process, and continuously constructs and updates time-series data. However, when there is no risk of failure, this data is primarily used for daily monitoring of the device's operating status and performance analysis.

[0059] The system also keeps a detailed record of the prediction results for this fault-free period. This record includes the input data for the fault prediction model—the corrected time series at the time—as well as the model's output, clearly indicating the risk of undetected equipment failures. This record is crucial for subsequent data analysis and model optimization. When scheduled maintenance or performance evaluation of the equipment is required, this data from the fault-free period helps operators understand the equipment's performance under normal conditions, enabling them to develop more effective maintenance plans and standards.

[0060] The self-service baggage check-in equipment fault prediction method provided in the above embodiment obtains operational data fragments collected by internal sensors and environmental data fragments collected by external sensors, combines them with timestamps to construct a complete time-series data sequence, and dynamically combines the equipment operating status with environmental conditions to ensure the comprehensiveness and relevance of the data. Subsequently, an interference detection model is used to identify and correct abnormal data, generating an interference-free corrected time data sequence, improving the continuity and accuracy of data input, and providing a high-quality data foundation for subsequent analysis. Based on the corrected time data sequence, the fault prediction model can accurately analyze the equipment operating status and environmental changes, detect potential fault risks in advance, and locate the specific fault location. Finally, by automatically generating maintenance requests, identifying risk locations and maintenance measures, it effectively reduces equipment downtime and maintenance costs, and improves the equipment's operating efficiency and reliability. The interaction of the above technical features enables the system to implement an efficient intelligent monitoring and prediction mechanism, significantly improving the stability of self-service baggage check-in equipment fault prediction in complex environments, and reducing the impact of external data interference on self-service baggage check-in equipment fault prediction.

[0061] In the above embodiment, the self-service baggage check-in equipment fault prediction system constructs a complete time series data sequence from the received operational data segments and environmental data segments, and uses an interference detection model to identify and correct abnormal data to obtain a corrected time series data sequence. The fault prediction model is used to predict faults for the self-service baggage check-in equipment. In actual applications, in some scenarios, some operational data segments or environmental data segments may experience transmission delays during transmission. Therefore, in some embodiments, the self-service baggage check-in equipment fault prediction system can first process the temporal nature of the data segments before constructing the time series data sequence. In addition, in terms of data correction, different types of data interference will have different effects on the data. Therefore, in some embodiments, the data correction process will select different correction methods based on the data interference type to improve the data correction effect and ensure data accuracy. Furthermore, when determining the fault risk location, in some embodiments, this can be accomplished by determining the sensor corresponding to the fault risk location, thereby more accurately locating the fault risk location.

[0062] like Figure 3FIG. 1 is another flow chart of a method for predicting failures of self-service baggage check-in equipment according to an embodiment of the present invention, which is described in detail as follows: 301. Obtain an operation data segment generated during the baggage check-in process and collected by an internal sensor of a self-service baggage check-in device. The operation data segment is a data segment with a preset internal sensor collection length.

[0063] Obtain environmental data segments around the self-service baggage check-in device collected by external sensors, where the environmental data segments are data segments with preset external sensor collection lengths.

[0064] Step 301 and Figure 2 Step 201 in the illustrated embodiment is similar, and reference may be made to the description in the relevant steps, which will not be repeated here.

[0065] 302. Extract a receiving timestamp and a collecting timestamp of each piece of received operation data and environment data, where the receiving timestamp is used to mark the receiving time of the operation data and the environment data, and the collecting timestamp is used to mark the collecting time of the operation data and the environment data.

[0066] Accurate timestamp extraction is crucial for the data processing process of the self-service baggage drop-off device failure prediction system. It is fundamental to ensuring the accuracy and timing consistency of subsequent data processing. As the system continuously receives operational data from various sensors within the self-service baggage drop-off device (such as weight sensors, position sensors, and speed sensors), as well as environmental data collected by external sensors (such as temperature sensors, humidity sensors, and air pressure sensors), the primary task is to accurately extract the receipt and collection timestamps from this data. For each piece of operational and environmental data, the system has a dedicated program module responsible for extracting timestamps. Obtaining the received timestamp is relatively straightforward. The system uses its internal high-precision clock module to record the instant the data enters. This time information is accurate to milliseconds or even microseconds, ensuring the precise timing of data arrival at the self-service baggage drop-off device failure prediction system.

[0067] Extracting the acquisition timestamp relies on the sensor's own time synchronization mechanism. On the sensor side, every data acquisition operation is accompanied by a time stamp. The sensor itself is equipped with a precise internal timing device, which is regularly synchronized and calibrated with the system clock to ensure time accuracy. When the sensor acquires data, its internal timing device records the instant as the acquisition timestamp.

[0068] As the system receives data, it quickly extracts the corresponding reception and acquisition timestamps from the data frame structure and stores them in association with the data content. This ensures that every piece of data, whether large amounts of operational or environmental data, is accurately timestamped, providing a solid time foundation for subsequent data processing and analysis. By extracting these two timestamps, the system clearly understands the data's transmission path and time, providing the necessary basis for determining whether data correction is necessary and maintaining data temporal consistency.

[0069] 303. Store the received operation data and environmental data in a data buffer of a corresponding sensor.

[0070] After extracting timestamps from the operational and environmental data, the self-service baggage drop-off device fault prediction system then stores this data in the corresponding sensor's data buffer. The data buffer is a temporary data storage area dedicated to each sensor within the system. It acts like a data "transfer station," efficiently managing and organizing data and providing organized data input for subsequent data processing.

[0071] During storage, data is arranged according to the first-in, first-out (FIFO) principle, with the first data received at the front of the buffer, and the last data received at the back. This storage method ensures the chronological order of the data, allowing subsequent processing and analysis to follow the order in which the data was collected, reproducing the true process of device operation and environmental changes.

[0072] 304. Subtract the data of the receiving timestamp from the data of the collecting timestamp to obtain a data delay value.

[0073] If the data delay value exceeds a preset threshold, the data of the received timestamp is subtracted from the data delay value to obtain a corrected timestamp.

[0074] After completing data storage, the self-service baggage check-in equipment failure prediction system enters the critical timestamp calculation and correction phase. This phase aims to calculate data delays by processing the received and collected timestamps, and correct the data when necessary to ensure proper alignment of operational and environmental data, providing an accurate data foundation for building a unified time series data set. The system first calculates the data delay by subtracting the received timestamp from the acquisition timestamp stored in the data buffer. This value clearly reflects the time interval between sensor acquisition and transmission to the system. In actual operation, the delay value for different data may vary due to various factors such as network transmission fluctuations and device performance.

[0075] To ensure data accuracy and timing consistency, the system pre-sets a reasonable threshold. This threshold is determined based on a comprehensive assessment of multiple factors, including the system's network environment, sensor performance, and data transmission requirements. When the calculated data delay exceeds the preset threshold, the system initiates a data correction process. Specifically, the received timestamp is subtracted from the data delay value to generate a corrected timestamp. This correction eliminates data timing distortions caused by transmission delays, ensuring that the corrected timestamps more accurately reflect the actual chronological order of data acquisition.

[0076] The process of obtaining the correction timestamp can eliminate the data timing disorder caused by transmission delay, so that the timing of operation data and environmental data can be correctly aligned, facilitating the construction of a unified time series data sequence.

[0077] 305. Sort the operation data and the environment data according to the correction timestamp to obtain a time series data sequence with a unified time sequence.

[0078] According to the time series data sequence, a plurality of time series data segments of preset lengths are obtained by division.

[0079] After completing timestamp correction, the self-service baggage check-in device failure prediction system entered the crucial phase of constructing and properly segmenting the time series data. This phase further organizes the preprocessed data and converts it into a format suitable for subsequent analysis models, laying the foundation for accurately detecting data interference and predicting device failures. The system sorts operational and environmental data based on the corrected timestamps. The corrected timestamps serve as the core basis for data sorting, ensuring that each piece of data is returned to its true chronological order. For example, during a baggage check-in process, the system receives operational data from a weight sensor. After timestamp correction, the actual time of collection is determined. Simultaneously, environmental data collected by the temperature sensor at the same time is also obtained. Using the corrected timestamps as a benchmark, the system accurately matches the weight data with the corresponding temperature data and then arranges this series of matched data in chronological order. As more data from various internal and external sensors continues to flow in, the system continues this sorting operation based on corrected timestamps, gradually building a complete multidimensional time series data set, or time series data sequence.

[0080] After constructing the time series data sequence, the system will then divide it to meet the data input format requirements of the subsequent analysis model. The system will cut the time series data sequence into several small segments, namely time series data segments, according to a pre-set fixed length.

[0081] Aligning and segmenting data into time series data segments offers significant advantages. It can significantly improve data processing efficiency. Compared to performing a global analysis of the entire massive time series data sequence, segmenting it into multiple smaller segments allows processing only one segment at a time, significantly reducing computational effort and memory usage.

[0082] 306. Input the time series data segment into an interference detection model to determine whether the time series data segment is interfered with.

[0083] Step 306 and Figure 2 Step 203 in the illustrated embodiment is similar, and reference may be made to the description in the relevant steps, which will not be repeated here.

[0084] 307. If interference occurs, compare the time series data segment with a preset normal data range to obtain an abnormal value that exceeds the normal data range.

[0085] Mark the time steps of the abnormal values in the time series data segment.

[0086] The type of interference that the time series data segment has been subjected to is determined based on the time step of the outlier.

[0087] When the self-service baggage check-in equipment fault prediction system determines that a certain time series data segment is interfered with through the interference detection model, it quickly enters the interference analysis and type judgment stage. This step is crucial for accurately understanding the source of the interference and taking effective corrective measures.

[0088] The system first compares the disturbed time series data fragments with the preset normal data range. The system's internal database pre-stores the reasonable value ranges of various sensor data under normal operating conditions.

[0089] Outlier labeling and interference type determination provide a basis for subsequent data correction, ensuring a more targeted correction process. After identifying outliers that fall outside the normal data range, the system further accurately labels the time steps of these outliers within the time series data segment. A time step refers to the time position corresponding to each data point in the time series data sequence. By labeling time steps, the system can clearly understand the specific time when the outlier occurred.

[0090] Based on the outlier's time step and related data characteristics, the system begins to determine the type of interference the time series data segment has experienced. Through comprehensive analysis of the outlier's time step and data characteristics, the system can accurately determine the type of interference, providing a strong basis for subsequent targeted data correction measures.

[0091] 308. If the abnormal value is a missing value caused by data collection loss, a correction value of the abnormal value is calculated by a linear interpolation formula based on the normal data before and after the abnormal value in the time series data segment. The linear interpolation calculation formula is: The abnormal value corresponding to the time step t is x t , the x t-1 and the x t+1 is the normal data before and after the abnormal value, the x t ' is the correction value of the abnormal value.

[0092] When the self-service baggage check-in equipment fault prediction system determines during interference analysis that an outlier is a missing value caused by data collection loss, it will quickly enable a linear interpolation algorithm to calculate a correction value for the outlier to restore the true state of the data to the greatest extent possible and ensure the integrity and accuracy of the data.

[0093] The linear interpolation formula assumes a linear trend in data change and estimates missing values by using the normal data before and after the outlier. Based on the preceding and following positions and time intervals, the value at the missing location is accurately estimated. In this way, the system can properly correct missing values caused by data acquisition loss, ensuring that the corrected data conforms to the overall variation pattern of the original data. This provides consistent and accurate data support for subsequent interference detection, fault prediction, and other processes, avoiding analytical bias caused by missing data.

[0094] 309. If the time step where the abnormal value is located deviates from the normal range by a high amplitude within a short period of time, it is determined to be a short-term spike abnormality.

[0095] The correction value is calculated using the sliding window average method for spike anomalies. The calculation formula of the sliding window average method is: described is the time before and after the time step t The data value of the time step, w is the set sliding window size, x t ' is the correction value of the abnormal value.

[0096] When the self-service baggage check-in equipment fault prediction system identifies that the time step where the outlier value deviates from the normal range for a short period of time with a high amplitude, that is, it is determined to be a short-term spike anomaly, the sliding window average method will be used to calculate the correction value of the outlier value to eliminate the adverse effects of the spike anomaly on the overall trend of the data and ensure the stability and reliability of the data.

[0097] After detecting a short-term spike anomaly, the system will immediately determine the time step where the abnormal value is located and set an appropriate sliding window around the time step. The size of the sliding window w is not fixed, but is dynamically adjusted by the system's intelligent algorithm based on the characteristics of different sensor data and the fluctuation of historical data. Taking the speed sensor data as an example, suppose that at a certain moment, the speed sensor outputs an abnormal value that is significantly higher than the normal range, and the system determines it as a short-term spike anomaly. If the speed data is relatively stable under normal circumstances and has small fluctuations, the system may set a smaller sliding window, such as w = 3, which includes the abnormal value time step t and the time steps before and after it. The correction calculation is performed based on the data of each time step. Assume that the time step of the outlier is t, and its predecessor The data value of the time step is back The data value of the time step is According to the sliding window average method, the calculation formula (Here w=3), correct the outliers.

[0098] The advantage of the sliding window average method lies in its ability to smooth out the impact of short-term, high-amplitude anomalies by comprehensively considering data from multiple adjacent time steps, allowing the corrected data to better reflect the true trends in equipment operation. In practical applications, whether it's a momentary overcurrent spike in motor current data or a brief high-temperature anomaly in a temperature sensor caused by sudden external heat interference, the system can use the sliding window average method to flexibly adjust the window size based on the data characteristics and accurately calculate the correction value. This provides a stable and reliable data foundation for subsequent data processing and analysis, effectively improving the accuracy of fault prediction.

[0099] In addition, when the data exhibits complex nonlinear changes and cannot be accurately estimated using simple interpolation or smoothing methods, and the system has sufficient historical data to train a high-precision forecast model, a data correction method based on the time series forecast model can also be used. The method is described as follows: We extract operational and environmental data similar to the current abnormal data fragment from historical data and train a time series forecasting model. This model is constructed using LSTM (Long Short-Term Memory) networks, which are suitable for complex nonlinear time series forecasting and can capture long-term dependencies.

[0100] The context data of the time step where the outlier is located (i.e., the normal data before and after the outlier) is input into the trained time series prediction model to predict the corrected value of the outlier.

[0101] In addition, when the distribution characteristics of outliers are highly similar to certain patterns in historical data, or when data collection is repeated multiple times and has a strong periodicity, such as when the equipment operates under specific conditions and environmental conditions repeatedly, a data correction method based on historical data supplementation is used. The specific method is described as follows: Filter out scenario data with similar characteristics to the current abnormal data fragment from historical data (such as operating data in the same time period, environmental data under similar conditions).

[0102] Based on historical data, directly match the normal value in the same scenario, or take the average of the normal values in multiple similar scenarios, and replace the abnormal value to obtain the corrected value.

[0103] 310. Obtain a corrected data segment of the time series data segment according to the correction value.

[0104] A corrected time data sequence is constructed based on the undisturbed time series data segments and the corrected data segments.

[0105] After completing the correction of outliers, the self-service baggage check-in equipment failure prediction system enters the key link of constructing a corrected data sequence. This step will integrate the undisturbed time series data fragments with the corrected time series data fragments to form a complete, accurate and interference-free corrected time data sequence, providing high-quality data input for subsequent equipment failure prediction. The system first thoroughly sorts and categorizes all time series data segments that have undergone interference detection and outlier correction. Time series data segments that have been determined to be undisturbed by the interference detection model are individually selected and placed in a temporary storage area. Furthermore, for time series data segments that have been interfered with and have undergone outlier correction, the system carefully checks the accuracy and completeness of their corrections to ensure that the corrected data conforms to the actual logic of device operation and environmental changes. After confirming the quality of all data segments, the system sequentially combines the undisturbed time series data segments with the corrected data segments, chronologically, to construct a continuous, complete corrected time data sequence.

[0106] In practice, we may encounter more complex situations, such as when there are multiple disturbed data segments scattered across the time series. In these cases, the system uses efficient sorting algorithms and data indexing techniques to quickly and accurately integrate all uninterrupted and corrected segments.

[0107] By constructing a corrected time data sequence, the system successfully eliminated interference factors and outliers in the data, allowing the final data sequence to truly and accurately reflect the actual operating status of the self-service baggage check-in equipment over a period of time and changes in the surrounding environment. At the same time, it lays a solid foundation for subsequent input into the fault prediction model to achieve accurate equipment failure prediction.

[0108] 311. Input the corrected time data sequence into a fault prediction model to obtain a fault prediction result of the self-service baggage check-in device.

[0109] Step 311 and Figure 2 Step 206 in the illustrated embodiment is similar, and reference may be made to the description in the relevant steps, which will not be repeated here.

[0110] 312. If the fault prediction result indicates that the self-service baggage check-in device will fail, determine the range of the operating data and the environmental data under normal operating conditions of the self-service baggage check-in device by analyzing the historical operating data and the environmental data.

[0111] By comprehensively and deeply analyzing historical operating and environmental data, the self-service baggage drop-off device fault prediction system can establish a precise and reliable data range system for the normal operation of the self-service baggage drop-off device, providing a solid foundation for subsequent identification of abnormal data. This range provides an important reference standard for subsequent accurate identification of abnormal data.

[0112] 313. If the fault prediction result indicates that the self-service baggage check-in device will fail, extract data segments in the corrected time data sequence that exceed the range of the operating data and the environmental data under the normal operating state to obtain abnormal data segments.

[0113] After determining the range of operating and environmental data for the self-service baggage drop-off device under normal operating conditions, the self-service baggage drop-off device failure prediction system then meticulously screens the corrected time data series, extracting data segments that fall outside the normal operating range, thus identifying abnormal data segments. This process is like pinpointing the problem within a massive amount of data, providing key clues for subsequent in-depth analysis of the fault cause and determining the location of the fault risk.

[0114] The system examines each data point in the corrected time data sequence, one by one, in chronological order. For example, consider weight sensor data within the operational data. Assume that the weight data range under normal operating conditions is determined to be between 5 kg and 40 kg. When the system processes the weight data within the corrected time data sequence and encounters a weight value of 45 kg at a specific moment, clearly outside the normal range, it immediately flags the data segment containing this anomalous weight value. This data segment doesn't simply contain the anomalous data point; rather, it selects data within a certain time span surrounding the anomalous point, based on pre-defined rules, to ensure that it captures all relevant information about the anomaly.

[0115] When extracting abnormal data segments, the system also uses data association analysis technology to determine whether there are correlations between anomalies in different types of sensor data. For example, if the extracted abnormal data segment not only finds that the weight sensor data is outside the normal range, but also finds that the position sensor data corresponding to this weight data shows that the luggage's position on the conveyor belt is abnormal, and the motor current also fluctuates significantly at this time, the system will comprehensively consider these related abnormal data and further confirm that the data segment is a significant abnormal data segment. Through this comprehensive and meticulous screening and extraction process, the system can accurately identify data segments that exceed the normal operating state data range from the corrected time data series, providing strong data support for the subsequent determination of the fault risk location.

[0116] 314. Determine a target sensor corresponding to the abnormal data segment according to the collection timestamp of the abnormal data segment.

[0117] After successfully extracting the abnormal data fragment, the self-service baggage check-in device fault prediction system enters the critical stage of identifying the target sensor. By analyzing the acquisition timestamp of the abnormal data fragment, the system can accurately locate the source of the abnormal data, that is, determine the target sensor, providing important clues for quickly and accurately identifying the fault risk location.

[0118] The acquisition timestamp is a unique and accurate time stamp recorded by each sensor when it collects data. It accurately reflects the moment the data was generated. Based on this precise record of the acquisition timestamp, the system can quickly identify the weight sensor as the key source of this abnormal data fragment, i.e., the target sensor.

[0119] 315. Determine the equipment component corresponding to the target sensor as the fault risk location.

[0120] After identifying the target sensor, the self-service baggage check-in equipment's fault prediction system then determines the fault risk location. By precisely matching the target sensor with the corresponding equipment component, the system can quickly and accurately identify the specific component at risk of failure, providing maintenance personnel with clear repair instructions and significantly improving the efficiency of equipment troubleshooting and repair. The system has a comprehensive database pre-built to map sensors to equipment components. This database details the equipment components monitored by each sensor, as well as the physical connections and functional relationships between them. By precisely locating the source of abnormal data and identifying the target sensor, and then matching the target sensor with the corresponding equipment component, the system can quickly and accurately identify fault risk locations among numerous equipment components, providing clear guidance for maintenance and avoiding blind investigations.

[0121] 316. Send a maintenance request for the fault risk location to a maintenance person.

[0122] 317. If the fault prediction result indicates that the self-service baggage check-in device will not fail, no response is made.

[0123] Step 316 and step 317 are the same as Figure 2 In the illustrated embodiment, step 207 is similar to step 208, and reference may be made to the descriptions in the relevant steps, which will not be repeated here.

[0124] The self-service baggage check-in equipment fault prediction method provided in the embodiment of the present application effectively reduces the impact of external interference on the fault prediction results through multi-step collaborative operation. First, the system obtains the operation and environmental data fragments and extracts the reception and collection timestamps, which provides a basis for judging the data transmission delay and ensuring the consistency of the timing. Then, the data is stored in the corresponding buffer and the delay value is calculated according to the timestamp, and the over-threshold delay is corrected so that the data is arranged in the real time order, laying a solid foundation for building a standardized time series data sequence. Then, the time series data sequence is divided into fragments and input into the interference detection model, the interference is identified and the outliers are found by comparing with the normal data range, the interference type is judged according to the outlier time step and data characteristics, and then linear interpolation or sliding window average method is used for correction of different interferences (such as data acquisition loss or short-term spike anomalies), to obtain the corrected data fragments and construct an interference-free corrected time data sequence, providing high-quality input for the fault prediction model and improving the prediction accuracy. When a device failure is predicted, the system determines the normal data range by analyzing historical data, extracts abnormal data fragments from the corrected time data sequence, identifies the target sensor based on the acquisition timestamp, and then determines the fault risk location and sends a maintenance request to the maintenance personnel, thereby achieving accurate positioning and efficient processing of the fault, comprehensively improving the reliability and practicality of the self-service baggage check-in equipment fault prediction system, and effectively ensuring the stable operation of the equipment.

[0125] The method provided in the above embodiment can be executed by a self-service baggage check-in equipment fault prediction system, which is composed of electronic devices. The following describes the electronic device in the embodiment of the present application from the perspective of hardware processing. Figure 4 , is a schematic diagram of a physical device structure of an electronic device in an embodiment of the present application.

[0126] It should be noted that Figure 4 The structure of the electronic device shown is only an example and should not limit the functions and scope of use of the embodiments of the present application.

[0127] like Figure 4As shown, the electronic device includes a central processing unit (CPU) 401, which can perform various appropriate actions and processes according to the program stored in the read-only memory (ROM) 402 or the program loaded from the storage part 408 into the random access memory (RAM) 403, such as the method described in the above embodiment. In the random access memory (RAM) 403, various programs and data required for system operation are also stored. The central processing unit (CPU) 401, the read-only memory (ROM) 402 and the random access memory (RAM) 403 are connected to each other via a bus 404. An input / output (I / O) interface 405 is also connected to the bus 404.

[0128] The following components are connected to the input / output (I / O) interface 405: an input section 406 including an audio input device, a push button switch, etc.; an output section 407 including a display, an audio output device, an indicator light, etc.; a storage section 408 including a hard disk, etc.; and a communication section 409 including a network interface card such as a LAN (Local Area Network) card or a modem. The communication section 409 performs communication processing via a network such as the Internet. A drive 410 is also connected to the input / output (I / O) interface 405 as needed. A removable medium 411, such as a magnetic disk, an optical disk, a magneto-optical disk, a semiconductor memory, etc., is installed in the drive 410 as needed so that a computer program read therefrom can be installed into the storage section 408 as needed.

[0129] In particular, according to an embodiment of the present application, the process described above with reference to the flowchart can be implemented as a computer software program. For example, an embodiment of the present application includes a computer program product, which includes a computer program carried on a computer-readable medium, and the computer program includes a computer program for executing the method shown in the flowchart. In such an embodiment, the computer program can be downloaded and installed from the network through the communication part 409, and / or installed from the removable medium 411. When the computer program is executed by the central processing unit (CPU) 401, the various functions defined in the present application are performed.

[0130] It should be noted that specific examples of computer-readable storage media may include, but are not limited to, an electrical connection having one or more conductors, a portable computer disk, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM), a flash memory, an optical fiber, a portable compact disc read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination thereof. In this application, a computer-readable storage medium may be any tangible medium containing or storing a program that can be used by or in conjunction with an instruction execution system, apparatus, or device.

[0131] The flowcharts and block diagrams in the accompanying drawings illustrate the possible architecture, functions and operations of the systems, methods and computer program products according to various embodiments of the present application. Each box in the flowchart or block diagram can represent a module, program segment, or part of the code, and the above-mentioned module, program segment, or part of the code contains one or more executable instructions for implementing the specified logical functions. It should also be noted that in some alternative implementations, the functions marked in the box can also occur in an order different from that marked in the accompanying drawings.

[0132] Specifically, the electronic device of this embodiment includes a processor and a memory, the memory is coupled to one or more processors, the memory is used to store computer program code, the computer program code includes computer instructions, and one or more processors call the computer instructions to enable the electronic device to execute the method provided by the above embodiment.

[0133] As another aspect, the present application further provides a computer-readable storage medium, which may be included in the electronic device described in the above embodiments, or may exist independently and not be incorporated into the electronic device. The above storage medium carries one or more computer programs, and when the one or more computer programs are executed by a processor of the electronic device, the electronic device implements the method provided in the above embodiments.

[0134] As described above, the above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit the same. Although the present invention has been described in detail with reference to the above embodiments, those skilled in the art should understand that the technical solutions described in the above embodiments can still be modified, or some of the technical features thereof can be replaced by equivalents. However, these modifications or replacements do not deviate the essence of the corresponding technical solutions from the scope of the technical solutions of the embodiments of the present invention.

[0135] As used in the above embodiments, the term “when…” may be interpreted to mean “if…” or “after…” or “in response to determining…” or “in response to detecting…”, depending on the context. Similarly, the phrases “upon determining…” or “if (stated condition or event) is detected” may be interpreted to mean “if determining…” or “in response to determining…” or “upon detecting (stated condition or event)” or “in response to detecting (stated condition or event)”, depending on the context.

[0136] Those skilled in the art will appreciate that all or part of the process steps in the above-described method embodiments can be implemented by a computer program instructing the relevant hardware. The program can be stored in a computer-readable storage medium, and when executed, the program can include the process steps in the above-described method embodiments. The aforementioned storage medium includes various media capable of storing program code, such as ROM or random access memory (RAM), magnetic disks, or optical disks.

Claims

1. A self-service baggage check-in equipment failure prediction method, characterized in that: include: Obtaining operational data segments generated during the baggage check-in process by internal sensors of the self-service baggage check-in device and environmental data segments collected by external sensors of the surrounding area, wherein the operational data segments are data segments of a preset length collected by the internal sensors, and the environmental data segments are data segments of a preset length collected by the external sensors; Constructing the operation data segments and the environment data segments into a time series data sequence; Dividing the time series data sequence into a plurality of time series data segments of preset lengths; Inputting the time series data segment into an interference detection model to determine whether the time series data segment is interfered with; If interference occurs, performing data correction on the time series data segment to obtain a corrected data segment; constructing a corrected time data sequence based on the undisturbed time series data segments and the corrected data segments; Inputting the corrected time data sequence into a fault prediction model to obtain a fault prediction result of the self-service baggage check-in device; If the fault prediction result indicates that the self-service baggage check-in device will fail, a maintenance request for the fault risk location is sent to maintenance personnel, where the fault risk location is the corresponding fault location when the self-service baggage check-in device is predicted to fail.

2. The method according to claim 1, characterized in that The training process of the interference detection model is: Extract features from the sample time series data segments to obtain operational data features and environmental data features. The operational data features include baggage weighing data, barcode scanning data, and motion status data. The environmental data features include temperature data, humidity data, and air pressure data. Construct a time series data segment training dataset containing multiple scenarios and multiple working conditions, wherein the time series data segment training dataset includes sample time series data segments and interference label annotations, and the interference labels include "interfered" labels and "undisturbed" labels; train the interference detection model based on the time series data segment training dataset.

3. The method according to claim 1, characterized in that The training process of the fault prediction model is as follows: Acquiring historical operating data, historical environmental data, and historical fault records collected during the operation of the self-service baggage check-in device, wherein the historical operating data includes baggage weighing data, barcode scanning data, and motion status data; the historical environmental data includes temperature data, humidity data, and air pressure data; and the historical fault records include operating data and environmental data at the time the fault occurred; Constructing a historical data training data set based on the historical operation data, the historical environmental data, and the historical fault records; The fault prediction model is trained based on the historical data training data set.

4. The method according to any one of claims 1 to 3, characterized in that If interference occurs, performing data correction on the time series data segment to obtain a corrected data segment specifically includes: If interference occurs, the time series data segment is compared with a preset normal data range to obtain abnormal values that exceed the normal data range; Marking the time steps of the outliers in the time series data segment; If the abnormal value is a missing value caused by data collection loss, a correction value of the abnormal value is calculated by a linear interpolation formula based on the normal data before and after the abnormal value in the time series data segment; A corrected data segment of the time series data segment is obtained according to the correction value.

5. The method according to claim 4, characterized in that After marking the time steps of the abnormal values in the time series data segment, the method further includes: If the time step where the abnormal value is located deviates from the normal range by a high amplitude within a short period of time, it is determined to be a short-term spike abnormality; The correction value is calculated using the sliding window average method for the short-term spike anomaly. The calculation formula of the sliding window average method is: described is the time before and after the time step t The data value of the time step, w is the set sliding window size, x t ' is the correction value of the abnormal value.

6. The method according to any one of claims 1 to 5, characterized in that The step of constructing the operation data segments and the environment data segments into a time series data sequence specifically includes: Extracting a reception timestamp and a collection timestamp of each piece of received operation data and environment data, wherein the reception timestamp is used to mark the reception time of the operation data and the environment data, and the collection timestamp is used to mark the collection time of the operation data and the environment data; Storing the received operating data and environmental data in a data buffer of a corresponding sensor; Subtract the data of the received timestamp from the data of the collected timestamp to obtain the value According to the delay value; If the data delay value exceeds a preset threshold, subtract the data of the received timestamp from the data delay value to obtain a corrected timestamp; According to the correction timestamp, the operation data and the environment data are sorted according to the correction timestamp to obtain a time series data sequence with unified time sequence.

7. The method according to any one of claims 1 to 6, characterized in that If the fault prediction result indicates that the self-service baggage check-in device is likely to fail, sending a maintenance request for the fault-risk location to a maintenance person specifically includes: Determine the range of the operating data and the environmental data when the self-service baggage check-in device is in normal operation by analyzing the historical operating data and the environmental data; If the fault prediction result indicates that the self-service baggage check-in device will fail, extracting data segments in the corrected time data sequence that exceed the range of the operating data and the environmental data under the normal operating state to obtain abnormal data segments; determining a target sensor corresponding to the abnormal data segment according to a collection timestamp of the abnormal data segment; The equipment component corresponding to the target sensor is determined as the fault risk location.

8. The self-service baggage check-in equipment failure prediction system is characterized by: including one or more processors and memory; The memory is coupled to the one or more processors, and the memory is used to store computer program code, where the computer program code includes computer instructions. The one or more processors call the computer instructions to enable the self-service baggage check-in equipment failure prediction system to execute the method according to any one of claims 1 to 7.

9. A computer-readable storage medium storing computer instructions, characterized in that: When the computer instructions are executed on a self-service baggage check-in equipment failure prediction system, the self-service baggage check-in equipment failure prediction system is caused to execute the method according to any one of claims 1 to 7.

10. A computer program product, characterized in that When the computer program product is run on a self-service baggage check-in equipment failure prediction system, the self-service baggage check-in equipment failure prediction system is caused to execute the method according to any one of claims 1 to 7.

Images