Data processing method and system for overhead passenger controller

Abstract

The invention relates to the technical field of data processing, in particular to a data processing method and system for an overhead passenger controller. The method comprises the following steps: acquiring position information data of the monkey vehicle and controller acquisition data of an aerial passenger controller where the monkey vehicle is located; determining an overhead man-riding device moving track based on the overhead man-riding device position information data, and performing three-dimensional modeling on the controller acquisition data according to the overhead man-riding device moving track to generate overhead man-riding device three-dimensional modeling data; and respectively analyzing an inertia disturbance characteristic, an inertia retardation delay characteristic and an intermittent load sudden change characteristic of the aerial cableway three-dimensional modeling data, and carrying out data characteristic distribution fusion to generate an aerial cableway controller behavior characteristic data set. By fusing three-dimensional modeling, dynamic feature analysis and image recognition, intelligent recognition and early warning control of the abnormal behavior of the aerial passenger controller are achieved, and the problems that a traditional system is single in perception, lagged in response and the like are solved.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of data processing, and particularly relates to a data processing method and system for an aerial passenger controller. BACKGROUND

[0002] Traditional controllers generally adopt PLC program logic control mode, which can complete basic start-stop control, running monitoring and alarm prompt functions, but has problems of response lag and insufficient precision in dealing with complex running states and abnormal behavior identification. In recent years, with the progress of sensor technology, embedded systems and data communication technology, controllers have gradually integrated inertial measurement units (IMU), multi-source signal acquisition modules and edge computing capabilities, so as to have real-time data processing and behavior identification capabilities. At the same time, based on the introduction of big data analysis and artificial intelligence algorithms, the controller has evolved from rule-driven to data-driven, which can realize accurate modeling and intelligent identification of inertial disturbance, load anomaly, signal mutation and other characteristics.

[0003] However, the traditional monitoring method has delay or even misses in identifying dynamic anomalies such as inertial disturbance and load mutation, which cannot meet the real-time response requirements in high-risk operating environments, and cannot integrate controller operating behavior and passenger action behavior data, resulting in misjudgment, missed judgment and other situations, especially in non-standard operation or sudden behavior scenarios. SUMMARY

[0004] Therefore, it is necessary to provide a data processing method and system for an aerial passenger controller to solve at least one of the above technical problems.

[0005] To achieve the above purpose, a data processing method for an aerial passenger controller, the method comprising the following steps: Step S1: acquiring monkey car position information data and controller acquisition data of the aerial passenger controller where the monkey car is located; confirming a monkey car running track based on the monkey car position information data, and generating monkey car three-dimensional modeling data by three-dimensional modeling of the controller acquisition data according to the monkey car running track; Step S2: analyzing inertial disturbance characteristics, inertial blocking delay characteristics and intermittent load mutation characteristics of the monkey car three-dimensional modeling data respectively, and performing data feature distribution fusion to generate monkey car controller behavior feature data set; Step S3: acquiring monkey car passenger shooting images; identifying personnel behavior feature data of the monkey car passenger shooting images, and performing behavior pairing and model construction on the monkey car controller behavior feature data and the personnel behavior feature data to generate a controller response judgment model; Step S4: The controller is judged for abnormal behavior by using the controller response judgment model, and when it is judged that there is abnormal behavior, a speed limiting, shutdown or alarm instruction is issued to the overhead passenger controller where the monkey car is located to perform corresponding early warning control operation.

[0006] The present application fuses monkey car position information, controller acquisition data and passenger image behavior information, establishes a multi-dimensional model containing motion characteristics and personnel interaction characteristics, so that the abnormal behavior recognition is more comprehensive and accurate. The common inertial disturbance, inertial lag and intermittent load mutation phenomenon in the running process of the monkey car are analyzed and modeled, which can effectively capture abnormal running signs and improve the adaptability of the system to nonlinear complex working conditions. Through behavior feature extraction in the passenger image, the behavior matching with the controller feature data is further performed, which makes up for the deficiency of traditional controller monitoring data in abnormal behavior interpretation and enhances the intuitiveness and logical reasonableness of the judgment. The controller response model based on multi-modal features can intelligently identify and quickly respond to abnormal behavior, improve the active prevention and control ability of the system and reduce the accident rate. Once it is judged that there is abnormal behavior, the system can automatically issue a speed limiting, shutdown or alarm instruction to realize the integration of "recognition-decision-control" closed loop response, significantly improve the real-time early warning ability and overall safety of the monkey car running system. The present application is especially suitable for the mine environment with dense personnel, complex terrain and frequent operation, and can improve the running stability of the overhead passenger system and the life safety protection ability of the operating personnel. Therefore, the present application realizes intelligent identification and early warning control of abnormal behavior of the overhead passenger controller by fusing three-dimensional modeling, dynamic feature analysis and image recognition, and solves the problems of single perception and response lag of the traditional system.

[0007] Preferably, step S1 comprises the following steps: Step S11: Obtain monkey car position information data and controller acquisition data of the overhead passenger controller where the monkey car is located; Step S12: Restore the coordinate trajectory of the monkey car position information data to generate three-dimensional running trajectory data of the monkey car; Step S13: Extract the controller timestamp of the controller acquisition data, and perform trajectory point mapping fusion combined with the three-dimensional running trajectory data of the monkey car to generate trajectory binding controller data set; Step S14: Perform spatial data interpolation and parameter normalization based on the trajectory binding controller data set to generate trajectory control parameter distribution data; perform spatial grid modeling based on the trajectory control parameter distribution data to generate three-dimensional modeling data of the monkey car.

[0008] The present application can accurately restore the three-dimensional running track of the monkey car by step-by-step processing of the monkey car position information and the controller collected data, especially realizing coordinate track restoration in S12, and ensuring that subsequent analysis is based on real and high-precision track data. Mapping and fusing the time-stamped controller data with the three-dimensional track points breaks the limitations of a single data source, forms a track-bound controller data set, and provides a solid data foundation for subsequent comprehensive modeling. Through spatial data interpolation and parameter normalization processing, data missing and differences are effectively filled, noise and abnormal points are eliminated, the continuity and consistency of the data are improved, and the stability of the model is enhanced. Based on the spatial grid modeling of the track control parameter distribution, the three-dimensional modeling data of the monkey car with spatial structure information is generated, which provides stereoscopic and multi-dimensional data support for monkey car behavior analysis and anomaly detection. The refined processing procedure helps to improve data quality and expression ability, promote the accuracy and real-time performance of subsequent steps such as inertial disturbance feature extraction and behavior feature fusion, and support intelligent early warning decision-making. This method fully considers the data characteristics in the space and time dimensions, provides a solid technical guarantee for the operation monitoring of the monkey car in the complex terrain and variable working conditions of the mine area, and improves the system robustness and safety.

[0009] Preferably, the step S14 of performing spatial grid modeling based on the track control parameter distribution data comprises: performing spatial discretization processing based on the track control parameter distribution data, dividing into a plurality of three-dimensional grid units, and generating initial spatial grid data; performing boundary condition identification on the initial spatial grid data, labeling the track-related grid units, and generating track-constrained grid data; calculating the spatial attribute parameters of each grid unit according to the track-constrained grid data, including position coordinates and motion state indicators, and generating grid attribute parameter data; performing topological connection between the grid units using the grid attribute parameter data, constructing a complete spatial grid model, and generating three-dimensional modeling data of the monkey car.

[0010] The application realizes fine-grained modeling of the monkey car running space by dividing the trajectory control parameter distribution data into multiple three-dimensional grid units, improves the accuracy and resolution of the space representation, and is beneficial to capturing the slight changes of complex trajectories. The boundary condition identification and the labeling of the trajectory-related grid units ensure that the modeling process is closely around the monkey car trajectory, enhances the expression ability of the space model to the trajectory constraints, and improves the pertinence of the abnormal behavior detection. By calculating the spatial position and motion state indicators of each grid unit, the spatial grid is given multi-dimensional attributes, realizing the deep description of the dynamic motion characteristics of the monkey car, and providing rich data support for subsequent behavior analysis. The topological connection between units is realized by using grid attribute parameters, forming a continuous and logically rigorous spatial grid model, improving the structural integrity and continuity of the model, and being beneficial to dynamic analysis and simulation. The meticulous grid modeling and attribute assignment make the monkey car three-dimensional data model have good visualization effect and data analysis performance, facilitating intuitive monitoring of the monkey car motion state and abnormal point distribution. The spatial grid modeling method has good flexibility, and can adjust the grid division and attribute calculation according to different trajectories and control parameters, adapting to the needs of complex mine environment and diversified operation conditions.

[0011] Preferably, the step S2 of analyzing the inertia disturbance characteristics of the three-dimensional modeling data of the monkey car comprises: extracting acceleration and angular velocity signals based on the three-dimensional modeling data of the monkey car, and generating inertia motion signal data; performing time domain filtering processing on the inertia motion signal data to remove high-frequency noise, and generating filtered inertia signal data; calculating the acceleration change rate and the angular velocity change rate by using the filtered inertia signal data, and obtaining inertia disturbance amplitude data; performing statistical analysis on the inertia disturbance amplitude data, extracting inertia disturbance characteristic indicators, and obtaining inertia disturbance characteristics.

[0012] The application extracts acceleration and angular velocity signals based on three-dimensional modeling data, captures real-time motion changes of the monkey car in space, and ensures the authenticity and timeliness of the inertial motion signals. High-frequency noise is removed by time domain filtering, reducing the influence of environmental interference and measurement error, ensuring that subsequent calculation and analysis are based on high-quality inertial signal data, and improving the accuracy of feature extraction. The rate of change of acceleration and angular velocity reflects the subtle changes in the motion state, and the amplitude fluctuations of the inertial disturbance are sensitive, which helps to identify abnormal motion behavior in time. Statistical analysis of the amplitude of the inertial disturbance is performed to obtain representative feature indicators, which facilitate quantitative description of the inertial disturbance characteristics of the monkey car and provide effective basis for subsequent anomaly detection and behavior identification. Based on the fine inertial disturbance feature data, the system can identify the inertial abnormality of the monkey car in operation earlier, improve the sensitivity and accuracy of the early warning, and ensure the safety of the monkey car operation. Inertial disturbance characteristics as important physical quantities, combined with other behavior feature data, construct a comprehensive judgment model, and improve the understanding and prediction ability of the system to complex operation state.

[0013] Preferably, the step S2 of analyzing the inertial retardation delay characteristics of the monkey car three-dimensional modeling data comprises: Acceleration and angular velocity signal extraction is performed on the monkey car three-dimensional modeling data to generate inertial motion signal data; Based on the inertial motion signal data, the time delay of the inertial retardation response is calculated to obtain inertial retardation delay data; Time domain and frequency domain analysis is performed on the inertial retardation delay data to extract delay characteristic parameters and generate inertial retardation characteristic parameter data; In combination with the inertial retardation characteristic parameter data, the influence of the retardation effect of the monkey car motion state is evaluated to obtain the inertial retardation delay characteristics.

[0014] The application can accurately reflect the dynamic lag characteristic of the monkey car caused by the blocking effect in the movement process by extracting the acceleration and angular velocity signals, calculating the time delay of the inertial blocking response, and providing key time scale information for the motion state analysis. By combining time domain and frequency domain analysis methods, the time sequence and frequency characteristics of the inertial blocking delay data are deeply mined, the recognition ability of the blocking phenomenon is enhanced, and the description accuracy and expression richness of the characteristic data are improved. By using the inertial blocking characteristic parameter data, the constraint and influence of the blocking effect on the motion state of the monkey car are comprehensively evaluated, the performance of the inertial blocking in the actual operation is revealed, and the motion abnormality is accurately judged. The inertial blocking delay characteristic as an important index of abnormal signal helps to early find the blocking abnormality and running lag problem in the monkey car movement, and enhances the response ability of the early warning system to abnormal behavior under complex working conditions. The analysis process is based on the time delay characteristics and frequency characteristics of the physical signal, gives the model stronger physical interpretation, and improves the credibility and stability of the behavior judgment model. The inertial blocking delay characteristic and other behavior characteristics such as inertial disturbance characteristic are combined to enrich the description dimension of the monkey car motion state, which helps to build a more comprehensive controller behavior characteristic data set and improve the model robustness.

[0015] Preferably, the step S2 of analyzing the intermittent load mutation characteristics of the three-dimensional modeling data of the monkey car comprises: extracting the load change time sequence based on the three-dimensional modeling data of the monkey car, and generating load time sequence data; performing segmented statistical analysis on the load time sequence data, identifying the intermittent nodes of load change, and generating intermittent load node data; calculating the load mutation amplitude at the intermittent load nodes, and generating load mutation amplitude data; detecting abnormal fluctuations based on the load mutation amplitude data, screening out significant intermittent load mutation characteristics, and obtaining intermittent load mutation characteristics.

[0016] The application can comprehensively and dynamically reflect the load fluctuation of the monkey car in the running process by extracting the load change time sequence based on three-dimensional modeling data and constructing continuous load time sequence data, facilitating subsequent behavior analysis. By using the segmented statistical analysis technology, the intermittent nodes in the load time sequence are identified, which helps to quickly locate the position of the mutation and improves the efficiency and accuracy of feature extraction. By calculating the mutation amplitude at each intermittent node, the mutation strength can be effectively measured, providing a quantitative index for judging whether it belongs to an abnormal state, enhancing the objectivity and data support capability of abnormal detection. Combined with the abnormal fluctuation detection method, the significant features are selected from the mutation amplitude, effectively avoiding misjudgment and omission, and improving the sensitive response ability of the early warning system to sudden load abnormalities. The intermittent load mutation is often highly related to the risk events such as hook sliding, personnel getting on and off the car, obstacle collision and the like in the actual operation. By extracting such features, the perception ability of potential abnormal behavior can be significantly improved. The intermittent load mutation features are combined with other behavior features such as inertial disturbance and inertial resistance to enhance the feature dimension and discrimination accuracy of the controller response model, and improve the adaptability of the model to complex scenes. The feature analysis method is suitable for sudden load fluctuations caused by terrain changes, non-standard personnel operation or equipment failure in the mining area, which helps to identify and respond to mutation risks in time, and improves the system safety and robustness.

[0017] Preferably, the personnel behavior feature data of the monkey car passenger photographed image is identified in step S3, and the monkey car controller behavior feature data and the personnel behavior feature data are behavior paired and model constructed. The monkey car passenger photographed image is subjected to image preprocessing to generate a definition-optimized image, wherein the image preprocessing includes image brightness enhancement, image geometric transformation and image resolution enhancement. Personnel posture key points are extracted based on the definition-optimized image data to generate personnel posture feature data. The personnel posture feature data is subjected to behavior classification to generate personnel behavior feature data. The personnel behavior feature data and the monkey car controller behavior feature data are time-synchronized and behavior-paired to generate behavior-paired data. A controller response judgment model is constructed based on the behavior-paired data to generate the controller response judgment model.

[0018] The application significantly improves the image definition and stability by performing image preprocessing operations such as brightness enhancement, geometric transformation and resolution enhancement on the captured image, and provides high-quality input images for subsequent pose recognition and behavior extraction. Based on the optimized image data, key points are extracted, and human body pose features are generated, which can fully reflect the body dynamic changes of personnel during the operation of the monkey car, and provide intuitive and structured data support for behavior judgment. By classifying the pose feature data, behaviors such as standing, sitting, getting on and off the car, and abnormal shaking are recognized, forming high-level personnel behavior semantic labels and enhancing the behavior understanding ability. Through the time synchronization and behavior pairing of personnel behavior features and controller behavior features, the relationship between personnel operation and equipment response can be accurately corresponded, and potential human-machine abnormal interaction behaviors can be identified. Based on the behavior pairing data, a controller response judgment model is trained and constructed, so that the model has the ability to infer the equipment response state from the image behavior and controller feature combination, and the intelligent level of abnormal discrimination is improved. The method can effectively identify abnormal control behaviors caused by improper operation of personnel, realize the linkage analysis of human factors and equipment behaviors, and adapt to the complex, variable and high-risk operating environment of mining areas. Through the closed-loop process of image preprocessing, feature extraction, behavior pairing and model construction, the full-link automation from raw visual data to intelligent discrimination model is realized, and the flexibility and promotion value of system deployment are improved.

[0019] Preferably, the time synchronization and behavior pairing of the personnel behavior feature data and the controller behavior feature data include: The personnel behavior feature data and the monkey car controller behavior feature data are respectively divided into time windows to generate multi-level time slice data; The multi-level time slice data is time-aligned to process the asynchronicity of the two groups of data in the time scale to generate a preliminary time alignment index; The complex action sequences in the personnel behavior feature data and the monkey car controller behavior feature data are decomposed based on a predefined micro-event dictionary to extract micro-event sequence data of extremely fine granularity, wherein each micro-event contains action type, duration and intensity features, and a behavior micro-event time sequence is generated; A three-dimensional similarity measurement index is designed in combination with the action type, intensity and time duration features of the micro-event, and a multi-dimensional behavior similarity matrix is constructed; According to the preliminary time alignment index and the multi-dimensional behavior similarity matrix, iterative weighted matching is performed to identify the best micro-event corresponding pairing path, and behavior pairing data is generated.

[0020] The application effectively solves the time scale difference between the two types of data in terms of sampling frequency, response delay, etc. by dividing the personnel behavior and controller behavior into time windows and constructing multi-level time slices, and combining the preliminary time alignment index, to realize high-precision time synchronization. The pre-defined micro-event dictionary is used to decompose the complex action sequence into micro-event sequences containing action type, duration, intensity, etc., which greatly improves the delicacy and structural degree of behavior expression, and is beneficial to the accurate capture of small abnormal actions. The three-dimensional behavior similarity index is designed in combination with the action type, duration and intensity, etc. and a multi-dimensional behavior similarity matrix is generated, so that the behavior matching not only depends on the time characteristics, but also has multiple judgment basis such as semantics, amplitude and duration, which improves the reliability and accuracy of behavior pairing. Through iterative weighted matching of the preliminary time index and the behavior similarity matrix, the influence of short-term errors and transient disturbances is effectively avoided, and a stable and logically consistent behavior pairing path is extracted, which improves the model's ability to mine real behavior linkage relationships. This method not only can identify the surface personnel action and controller response relationship, but also can deeply mine the potential micro-behavior linkage mode, and provide a logically rigorous pairing basis for the controller response model driven by behavior. Through in-depth matching at the micro-event level, the subtle corresponding relationship between 'normal operation-abnormal response' or 'abnormal action-abnormal response' can be identified, which greatly enhances the system's perception and recognition ability of complex abnormal linkage events. The multi-dimensional similarity matching framework and micro-event structure constructed by the method can be flexibly expanded and applied to different types of behavior data matching scenarios, and has good algorithm universality and cross-application migration ability.

[0021] Preferably, step S4 comprises the following steps: Step S41: performing abnormal feature matching on the real-time behavior data of the controller based on the controller response judgment model to generate abnormal behavior discrimination result data; Step S42: performing risk level evaluation on the abnormal behavior discrimination result data to generate abnormal behavior risk level data; Step S43: determining the corresponding warning control instruction type according to the abnormal behavior risk level data to generate warning control instruction data, wherein the warning control instruction type includes speed limiting, shutdown or alarm; Step S44: sending an instruction to the overhead passenger controller where the monkey car is located through the warning control instruction data to trigger the corresponding warning control operation.

[0022] The application can quickly and automatically identify potential abnormal control behavior by matching the real-time collected controller behavior data with the abnormal feature of the response judgment model of the controller, avoiding manual delay judgment of the emergency and improving the system response efficiency. The risk level evaluation mechanism is introduced, which can quantize the risk level according to the severity, influence range and duration of the abnormal behavior, realize the transition from 'abnormal or not' to 'abnormal degree', and provide decision basis for grading response. Different types of control instructions (speed limit, shutdown, alarm) are generated according to different risk levels, realizing differentiated processing strategy, ensuring that high-risk abnormalities are handled decisively, while low-risk abnormalities are mainly controlled flexibly or prompted, ensuring the balance of operation efficiency and safety. Through real-time issuing and executing of control instructions such as speed limit, shutdown and alarm, the 'identification-evaluation-response' closed-loop mechanism is effectively established, ensuring that the system can trigger response immediately once abnormal behavior is identified, and significantly improving the timeliness and automation level of overall safety control. The control strategy is automatically decided according to model judgment and risk evaluation, reducing the dependence on manual experience judgment, and effectively avoiding safety accidents such as equipment collision, personnel injury or system downtime caused by misoperation or delayed judgment. The early warning control instruction can be flexibly configured according to actual needs (such as adding 'deceleration' and 'early warning broadcast' types), which has good expansibility and adaptability, and is suitable for complex operation scenes of different mining areas and different equipment models. The whole process data of each abnormal identification, risk level evaluation and control instruction execution are structured and saved, forming a traceable data chain, which is convenient for subsequent safety analysis, model optimization and operation strategy adjustment.

[0023] In the present specification, a data processing system for an aerial passenger controller is provided for executing the above-mentioned data processing method for an aerial passenger controller, which comprises: A three-dimensional modeling module is configured to acquire monkey car position information data and controller acquisition data of an aerial passenger controller where the monkey car is located; confirm a monkey car running track based on the monkey car position information data, and generate monkey car three-dimensional modeling data by performing three-dimensional modeling on the controller acquisition data according to the monkey car running track; A control behavior feature module is configured to analyze inertia disturbance features, inertia blocking delay features and intermittent load mutation features of the monkey car three-dimensional modeling data respectively, and perform data feature distribution fusion to generate a monkey car controller behavior feature data set; A user behavior feature module is configured to acquire monkey car passenger shooting images; identify personnel behavior feature data of the monkey car passenger shooting images, and perform behavior pairing and model construction on the monkey car controller behavior feature data and the personnel behavior feature data to generate a controller response judgment model; The pre-warning control module is used for judging the abnormal behavior of the controller by using the controller response judgment model, and when it is judged that there is an abnormal behavior, a speed limiting, shutdown or alarm instruction is issued to the overhead passenger controller where the monkey car is located to perform corresponding pre-warning control operation.

[0024] The application has the advantages that the running track of the monkey car is fused and modeled with the data collected by the controller by the three-dimensional modeling module, a high spatial precision and dynamic real-time three-dimensional running model of the monkey car is established, and a real and structured spatial semantic basis is provided for subsequent behavior analysis. The control behavior characteristic module deeply analyzes dynamic characteristics such as inertial disturbance, inertial resistance delay and intermittent load mutation, and fuses the data distribution, effectively describes the running state of the controller under complex working conditions, and greatly improves the recognizability and distinguishability of abnormal control behavior. The user behavior characteristic module pairs and models the image behavior of personnel and the behavior of the controller, innovatively introduces the modeling of the influence of human behavior on the response of the controller, opens up the abnormal analysis path of man-machine cooperation, and significantly enhances the intelligent identification ability of the system to operation risks. The pre-warning control module judges the behavior by the controller response judgment model, and automatically executes control instructions such as speed limiting, shutdown or alarm according to the judgment result, forms a real-time, automatic and accurate response closed-loop pre-warning system, and improves the overall safety guarantee level. The system makes decisions and responses based on data driving, reduces the dependence on manual experience judgment, and reduces the risk of control abnormalities and accidents caused by human factors such as operation errors and reaction delays. The modules are decoupled and clear, support flexible expansion and replacement, such as different types of image recognition algorithms, abnormal detection models and control strategies, meet the safety control needs of various complex scenes such as mines and industrial transportation. Therefore, the application realizes intelligent identification and pre-warning control of abnormal behavior of the overhead passenger controller by fusing three-dimensional modeling, dynamic characteristic analysis and image recognition, and solves the problems of single perception and response lag of the traditional system. BRIEF DESCRIPTION OF DRAWINGS

[0025] Figure 1 It is a step flowchart for a data processing method for an overhead passenger controller; Figure 2 It is a step flowchart for a data processing method for an overhead passenger controller; Figure 1 It is a detailed implementation step flowchart of step S1 in the application; Figure 3 It is a detailed implementation step flowchart of step S1 in the application; Figure 1 It is a detailed implementation step flowchart of step S4 in the application; The implementation, functional characteristics and advantages of the application will be further described with reference to the embodiments and the accompanying drawings. DETAILED DESCRIPTION

[0026] The technical method of the present application will be described clearly and completely below in conjunction with the drawings. Obviously, the described embodiments are part of the embodiments of the present application, rather than all the embodiments. Based on the embodiments in the present application, all other embodiments obtained by those skilled in the art without creative work fall within the scope of protection of the present application.

[0027] In addition, the drawings are only schematic illustrations of the present application and are not necessarily drawn to scale. Identical reference numerals in the drawings represent identical or similar parts, and thus repeated descriptions thereof will be omitted. Some of the block diagrams shown in the drawings are functional entities, which do not necessarily have to correspond to physically or logically independent entities. The functional entities can be implemented in the form of software, or in one or more hardware modules or integrated circuits, or in different network and / or processor methods and / or microcontroller methods.

[0028] It should be understood that although the terms "first", "second" and the like can be used herein to describe various elements, these elements should not be limited by these terms. These terms are only used to distinguish one element from another. For example, without departing from the scope of the example embodiments, a first element can be called a second element, and similarly a second element can be called a first element. The term "and / or" as used herein includes any and all combinations of one or more of the associated listed items.

[0029] To achieve the above-mentioned purpose, please refer to Figures 1 to 3 A data processing method for an aerial passenger controller, the method comprising the following steps: Step S1: Obtain gondola position information data and controller acquisition data of the aerial passenger controller where the gondola is located; confirm the gondola running track based on the gondola position information data, and generate gondola three-dimensional modeling data by three-dimensional modeling of the controller acquisition data according to the gondola running track; Step S2: Analyze the inertia disturbance characteristics, inertia blocking delay characteristics and intermittent load mutation characteristics of the gondola three-dimensional modeling data respectively, and perform data feature distribution fusion to generate gondola controller behavior feature data set; Step S3: Obtain gondola passenger shooting images; identify personnel behavior feature data of the gondola passenger shooting images, and perform behavior pairing and model construction of the gondola controller behavior feature data and the personnel behavior feature data to generate a controller response judgment model; Step S4: Use the controller response judgment model to judge the abnormal behavior of the controller, and when it is judged that there is an abnormal behavior, issue a speed limiting, shutdown or alarm instruction to the aerial passenger controller where the gondola is located to perform corresponding early warning control operation.

[0030] In the embodiment of the present application, the positioning device (such as UWB, inertial navigation or laser ranging equipment) erected in the mine roadway obtains real-time position information data of the monkey car, including dynamic coordinate data in the lateral, longitudinal and vertical directions. The sampling frequency can be set to more than 20 times per second to ensure the continuity and accuracy of the trajectory. At the same time, the controller acquisition data of the overhead man-riding controller where the monkey car is located is synchronously obtained. The controller data includes traction current, voltage change, brake instruction response time, equipment running state code, etc. Based on the position information data of the monkey car, a running trajectory model of the monkey car in the three-dimensional roadway environment is constructed. The controller acquisition data is aligned with the position information according to the time stamp, combined with the roadway section and slope information, and three-dimensional modeling calculation is performed. Finally, three-dimensional modeling data of the monkey car is generated for subsequent behavior feature extraction. The three-dimensional modeling data constructed in step S1 is analyzed, and the following three types of controller behavior features are extracted: identify the speed fluctuation of the monkey car under non-braking conditions, such as sudden acceleration or deceleration phenomenon, calculate the acceleration peak value, change rate and duration, etc. Analyze the delay of the actual action of the monkey car after the controller sends a running or stopping instruction. Extract indicators such as instruction response time difference and inertial sliding distance. Through the fluctuation of current, voltage and traction power, it is judged whether the monkey car appears sudden loading or unloading phenomenon. If the power change amplitude per unit time exceeds the preset value, it is determined that the load is suddenly changed. The above three types of features are distributed according to the spatial position and time sequence, and are fused to form a structured controller behavior feature data set containing feature point coordinates, feature type, feature value and time stamp. Through the image acquisition device installed in the monkey car compartment or around the monkey car, the monkey car passenger image data is obtained. The image sampling frequency is recommended to be no less than 10 frames per second to ensure the continuity of motion capture. Use video behavior recognition algorithm to process image data, identify passenger behavior features, including pulling, jumping, stepping, abnormal posture (such as standing, sudden standing up) and other behaviors. Extract corresponding time, spatial position and action category information for each type of behavior to generate personnel behavior feature data. The controller behavior feature data obtained in step S2 and the personnel behavior feature data are matched according to the time axis and spatial position to construct a relationship model between personnel behavior and controller response. Use machine learning models (such as random forest or long short-term memory neural network) to train a controller response judgment model to identify which personnel behaviors are likely to cause abnormal device response or have potential risks. The system runs the controller response judgment model in real time to jointly analyze the controller behavior features and personnel behavior features currently collected. If the model judgment result shows that the current behavior combination is highly similar to the historical abnormal pattern, it is judged as an abnormal behavior.When the abnormal behavior discrimination result appears, the system automatically selects the corresponding control response strategy according to the abnormal level: if it is a slight abnormality, such as short-time sliding delay, the system issues a speed limit instruction to reduce the speed of the monkey car; if it is a moderate abnormality, such as load mutation accompanied by abnormal action of passengers, the system issues a shutdown instruction and cuts off the main power supply; if it is a serious abnormality, such as inertia disturbance and passengers' violent behavior, the system immediately triggers an audible and light alarm, and uploads the data to the mine dispatching system or the central control room. All control responses are issued in real time through the communication bus of the overhead passenger controller of the monkey car, and are recorded in the log for subsequent tracing and optimizing the model.

[0031] As an example of the present application, reference is made to Fig. 1, which shows a monkey car system according to the present application. Figure 2 In this example, the step S1 includes: Step S11: Obtain the monkey car position information data and the controller acquisition data of the overhead passenger controller where the monkey car is located; Step S12: Restore the coordinate trajectory of the monkey car position information data to generate three-dimensional running trajectory data of the monkey car; Step S13: Extract the controller timestamp of the controller acquisition data, and combine the three-dimensional running trajectory data of the monkey car to perform trajectory point mapping fusion to generate a trajectory binding controller data set; Step S14: Perform spatial data interpolation and parameter normalization based on the trajectory binding controller data set to generate trajectory control parameter distribution data; and perform spatial grid modeling based on the trajectory control parameter distribution data to generate three-dimensional modeling data of the monkey car.

[0032] In the embodiment of the present application, by deploying a high-precision positioning system (such as a UWB positioning module, an inertial measurement unit IMU or a laser radar) in the monkey car operating environment, the position information data of the monkey car in the tunnel is collected in real time. The data contains coordinate values in three dimensions of lateral, longitudinal and height, as well as timestamp information, which is used for subsequent trajectory restoration. At the same time, the controller acquisition data is synchronously extracted from the overhead man-trip controller where the monkey car is located. The controller acquisition data includes: controller running state code (such as start, brake, stop); electrical parameters such as current, voltage, traction power; control instruction execution timestamp; sensor input data (such as inclination angle, speed, load change, etc.), through a unified time synchronization mechanism, the controller data and the positioning data are time-aligned to ensure data consistency and fusion. The position information data of the monkey car obtained by step S11 is used to reconstruct the trajectory of the coordinate points. An interpolation smoothing algorithm (such as B-spline interpolation) is used to eliminate jitter noise and restore the continuous running path of the monkey car in the tunnel. The trajectory data is organized in time sequence, and each trajectory point contains position coordinates (lateral, longitudinal, height), speed change trend and acquisition time. Finally, the three-dimensional running trajectory data of the monkey car is output, which is used for subsequent space fusion modeling. The timestamp of each control operation is extracted from the controller acquisition data, and the controller action data is bound to the corresponding trajectory point by matching with the time axis of the three-dimensional trajectory data. In the binding process, a time window matching method is used to ensure that each trajectory point can be associated with the controller state in the nearest or nearest time period. After fusion, each trajectory point not only contains spatial position information, but also embeds running parameters such as control current, load state and speed instruction of the controller at that time. Finally, the trajectory binding controller data set is generated, and the data structure is: trajectory point position information (three-dimensional coordinates), control parameters (current, voltage, load, speed, etc.), corresponding timestamp. Based on the trajectory binding controller data set, the following two processing steps are performed: interpolating the control parameters between the trajectory points to make up for the sparseness caused by data sampling interval. Normalize the multi-dimensional control parameters such as current, voltage and speed to the same scale to facilitate spatial distribution modeling. Project the entire trajectory data into a spatial grid, for example, based on a grid unit with a spacing of 1 meter, count and aggregate the average control parameters in each grid. A spatial grid model containing spatial coordinates and control parameter characteristic values is constructed. The output monkey car three-dimensional modeling data is a structured grid data set, each cell of which contains: three-dimensional spatial position index, average speed, current, load and other parameters corresponding to the position, control response delay, inertia trend and other implicit characteristic indicators. The modeling result can provide spatial basic data support for subsequent anomaly identification, path optimization and equipment state monitoring.

[0033] Preferably, the step S14 of performing spatial grid modeling based on the trajectory control parameter distribution data comprises: Discretize the space based on the trajectory control parameter distribution data, divide into several three-dimensional grid units, generate initial space grid data; Identify the boundary conditions of the initial space grid data, label the trajectory related grid units, generate trajectory constraint grid data; Calculate the spatial attribute parameters of each grid unit according to the trajectory constraint grid data, including position coordinates and motion state indicators, generate grid attribute parameter data; Connect the grid units using grid attribute parameter data, build a complete space grid model, generate monkey car three-dimensional modeling data.

[0034] In the embodiment of the application, the entire space region is discretely divided according to the geometric size and trajectory density characteristics of the monkey car running area, and the three-dimensional space is divided into multiple cubic grid units by setting a reasonable grid size (such as 1m*1m*1m). In the process of space discretization, the trajectory control parameter distribution data is mapped into the corresponding grid unit according to the corresponding three-dimensional coordinate position, forming the initial space grid data. Each grid unit records the speed, current, load and other normalized control parameter values carried by the trajectory points in the region. The initial space grid data is subjected to trajectory boundary condition identification, including: detecting all grid units through which the trajectory path passes; marking the grid containing the monkey car trajectory point as "trajectory related grid unit"; performing buffer expansion processing (such as expanding 1 grid unit) on the surrounding grid, which is used to support subsequent topological modeling and abnormal analysis. Through this process, trajectory constraint grid data is generated, which clearly indicates which grids belong to the effective trajectory area, which helps to limit the analysis range and improve the modeling efficiency. Based on the trajectory constraint grid data, the key spatial attributes of each grid unit are calculated and counted, including: center position coordinates (such as X, Y, Z values of the geometric center of the grid); control parameter mean value and distribution characteristics (such as average speed, current, control delay of all trajectory points in the grid); motion state indicators, such as the speed variation trend, acceleration interval, inertia disturbance amplitude of the trajectory points in the region; trajectory density or traffic frequency indicators for evaluating the importance or activity of the grid. Finally, the grid attribute parameter data of each unit is output, which provides the basis for subsequent topological modeling and abnormal identification. When building a complete three-dimensional model, based on the grid attribute parameter data, the topological structure between the space grids is established in the following way: an adjacency matrix is constructed according to the grid adjacency relationship (such as six-neighborhood, twenty-six-neighborhood); the connection state of each grid unit and its adjacent grid is recorded; if the trajectory attributes of adjacent grid units are continuous (such as consistent speed direction and control response), a connected relationship is established; a space graph structure is constructed, forming a topological grid graph with path continuity and control behavior coherence. The finally generated monkey car three-dimensional modeling data is a structured grid graph model, which contains not only trajectory paths and control parameters, but also spatial continuity and topological logical relationship.

[0035] Preferably, the inertia disturbance feature of the monkey car three-dimensional modeling data in step S2 includes: Based on the monkey car three-dimensional modeling data, acceleration and angular velocity signals are extracted to generate inertia motion signal data; The inertia motion signal data is subjected to time domain filtering processing to remove high frequency noise to generate filtered inertia signal data; The filtered inertia signal data is used to calculate acceleration change rate and angular velocity change rate to obtain inertia disturbance amplitude data; The inertia disturbance amplitude data is subjected to statistical analysis to extract inertia disturbance feature indexes to obtain inertia disturbance features.

[0036] In the embodiment of the present application, the speed, acceleration and attitude change information containing time sequence are extracted from the constructed monkey car three-dimensional modeling data, and the sensor data recorded in the controller acquisition data are combined, such as the three-axis acceleration values output by the acceleration sensor, the angular velocity signals (pitch angular velocity, roll angular velocity and yaw angular velocity) provided by the gyroscope. These information is integrated to generate inertia motion signal data containing time stamp, three-dimensional coordinates, linear acceleration and angular velocity, which is used as the basis input for subsequent analysis. Since there is certain noise and mechanical jitter in the acquisition data, in order to improve the accuracy of inertia feature analysis, the inertia motion signal data is subjected to time domain filtering processing: moving average, low-pass filtering or Kalman filtering method is used to suppress high frequency jitter and mutation error; a reasonable sliding window width (such as 0.2-0.5 seconds) is set to smooth the inertia change curve; the processed smooth acceleration curve and angular velocity curve are output to generate filtered inertia signal data. Based on the filtered signal data, the dynamic change rate of inertia disturbance is calculated to quantify the impact features and mutation degree in the monkey car motion, mainly including: the difference change of acceleration in a continuous time period, which is used to identify instantaneous acceleration or sudden stop behavior; identifying rotation impact, attitude mutation and other actions, especially suitable for uphill or curve sections; the above change rate is normalized to unify the unit standard, and the maximum value, average value, variance and other statistical values are calculated to output inertia disturbance amplitude data. The disturbance amplitude data is further analyzed to extract key indexes reflecting the running stability and safety of the monkey car, including but not limited to: identifying whether the inertia impact exceeding the set threshold value appears; judging whether the inertia disturbance is instantaneous or continuous; the number of times higher than the set disturbance threshold value per unit time; statistics of the direction (up and down / left and right / forward and backward) where the inertia disturbance mainly occurs; the fluctuation energy feature reflects the energy density of the overall inertia disturbance, which is used to evaluate the ride comfort or equipment loss risk, and the final output result is the inertia disturbance feature.

[0037] Preferably, the inertia disturbance feature of the monkey car three-dimensional modeling data in step S2 includes: Acceleration and angular velocity signal extraction is performed on the three-dimensional modeling data of the monkey car to generate inertial motion signal data; Based on the inertial motion signal data, the time delay of the inertial retardation response is calculated to obtain inertial retardation delay data; Time domain and frequency domain analysis is performed on the inertial retardation delay data to extract delay characteristic parameters to generate inertial retardation characteristic parameter data; Combined with the inertial retardation characteristic parameter data, the retardation effect influence of the monkey car motion state is evaluated to obtain the inertial retardation delay characteristic.

[0038] In the embodiment of the present application, the three-dimensional modeling data of the monkey car is processed by using an inertial signal solving tool. The three-dimensional modeling data is derived from a sequence of coordinate points on the overhead running path of the monkey car and corresponding controller sampling data. Each trajectory point contains spatial position (three-dimensional coordinates x, y, z) and timestamp information. The position change rate in the adjacent time period is calculated by calculating the velocity vector of each trajectory point. The formula v(t) = [x(t+1) -x(t)] / Δt is used, where Δt is the sampling time interval, which is uniformly set to 0.1 seconds. Then, the change of the velocity vector over time is differentiated, that is, the three-axis acceleration values Ax, Ay, and Az are obtained by using a(t) = [v(t+1) -v(t)] / Δt. The angular velocity is obtained based on the spatial direction change between the trajectory points. The change rate of the vector angle formed by the adjacent three trajectory points is calculated, and the Rodrigues rotation formula is used to convert it into Euler angles, and then the derivative is taken to obtain the angular velocity Gx, Gy, and Gz around the three coordinate axes. Finally, the above acceleration and angular velocity time series data are recorded as inertial motion signal data in a unified structured format. Each time point contains six inertial dimension signals Ax, Ay, Az, Gx, Gy, and Gz, and the frequency is uniformly set to 10 Hz, covering the entire trajectory. To measure the inertial response delay of the monkey car to the control command, the command sequence in the controller acquisition data needs to be classified and recognized first. The controller command includes three types of “acceleration command”, “deceleration command” and “stop command”, each with a timestamp accurate to milliseconds, for example, the “acceleration command” is recorded at T0. The controller command time is aligned with the inertial motion signal time series, and the response starting point of each type of inertial indicator is detected one by one after the command is issued. Taking acceleration Ax as an example, the change trend of Ax is detected in a window after T0. When the growth amplitude of Ax is not less than 0.2 m / s² and the single difference is greater than 0.1 m / s² at three consecutive time points (for example, T1, T2, T3), T1 is determined as the response starting point. Then the delay ΔT = T1-T0. This method is applied to each type of command respectively, and multiple ΔT values are extracted in each response period to form a complete inertial retardation delay data set. The above processing flow uses fixed response threshold settings: the acceleration threshold is 0.2 m / s², the angular velocity threshold is 10 deg / s, the minimum effective response window length is 3 time points (0.3 seconds), and the maximum observation time is 3 seconds. Finally, the delay result between each control command and its response time is recorded as a record and stored as inertial retardation delay data. The time domain analysis of the delay data is realized by calculating the average response time, the maximum delay time, the standard deviation of the response time, and other statistical characteristic indicators.The delay data set corresponding to each type of control command is counted respectively, such as for all the delay results of the acceleration command, the average value (e.g. 1.45 seconds), the maximum value (e.g. 2.75 seconds), the minimum value (e.g. 0.65 seconds) and the standard deviation (e.g. 0.58 seconds) are calculated. At the same time, the abnormal delay rate, that is, the ratio of the number of samples whose delay exceeds 2.5 seconds to the total number of samples, is counted. The threshold is set to 10%. The frequency domain analysis uses the fast Fourier transform (FFT) algorithm to perform frequency spectrum analysis on the acceleration signal sequence. The frequency analysis range is set to 0.1 Hz to 5 Hz, corresponding to the wavelength range of 2 seconds to 10 seconds. The main frequency component and its amplitude in each inertial signal channel (Ax, Ay, Az) are taken as the frequency domain features, such as the main frequency of 0.45 Hz and the amplitude of -35 dB. All the extracted feature parameters include but are not limited to average delay, maximum delay, frequency spectrum main frequency, amplitude peak value, abnormal delay proportion, etc. They are uniformly encoded and named fields to form complete inertial blocking feature parameter data. Combined with the inertial blocking feature parameter data obtained in the above, the blocking response state of the monkey car in different running stages is further analyzed and evaluated. By mapping the feature parameters back to the original trajectory node positions, the average delay changes corresponding to different positions or sections (such as tunnel entrances, turning points and slope starting sections) are compared, and high blocking response areas are marked. The evaluation indicators include the following contents: first, the trajectory segment blocking mean change rate is calculated, which is ΔT2-ΔT1, where ΔT2 and ΔT1 are the average delay values of two adjacent trajectory segments; second, the number of maximum blocking responses and the abnormal response proportion in each trajectory segment are recorded. If the abnormal response rate in a trajectory segment exceeds 20%, the segment is marked as a high blocking area; third, by comparing the differences in lagging features of the same control command in different control cycles, the correlation between structural blocking and control lag is analyzed. Finally, the results in the above evaluation process are uniformly encoded and output to generate inertial blocking delay features, which are used for subsequent matching with passenger behavior data or training of controller response models.

[0039] Preferably, the step S2 of analyzing the intermittent load mutation characteristics of the three-dimensional modeling data of the monkey car comprises: extracting a load change time sequence based on the three-dimensional modeling data of the monkey car to generate load time sequence data; segmenting and statistically analyzing the load time sequence data to identify intermittent nodes of load change and generate intermittent load node data; calculating the load mutation amplitude at the intermittent load nodes to generate load mutation amplitude data; detecting abnormal fluctuations based on the load mutation amplitude data to filter out significant intermittent load mutation characteristics to obtain intermittent load mutation characteristics.

[0040] In the embodiment of the present application, the real-time load record of the monkey car in the whole running process is extracted by calling the data channel bound with the load monitoring module in the controller acquisition system. The unit of the load data is kilogram (kg), the sampling frequency is set to 1 Hz, that is, one data is recorded per second, and each data contains a collection timestamp and a corresponding load value. Each trajectory node in the three-dimensional modeling data contains a controller number and a time identifier, the load data is synchronously matched with the trajectory time axis, a load change time sequence completely aligned with the trajectory of the monkey car is constructed, and the load time sequence data is uniformly named, and the structure format is: {t1, L1}, {t2, L2},..., {t n , n}, where t is the time point, and L is the actual load value at that time. Ensure that the data sequence covers the entire running section of the trolley from the starting point to the stopping point, and the integrity requires at least 95% of the continuous time period data to be valid. After the load time series data preprocessing, a fixed sliding window difference algorithm is used to identify the existing load mutation area. Set the sliding window width to 5 seconds (i.e. 5 sampling points), and the load change amplitude in each window is counted. Define the change amplitude ΔL as the difference between the load value at the end of the window and the load value at the beginning. If the absolute value of ΔL is greater than the set threshold of 10 kg, it is determined that the window contains intermittent change nodes. To locate the specific position of the intermittent node, further point-to-point difference is performed inside the window, and the load difference between the two consecutive time points is calculated. If the load change between any two points is greater than 15 kg and there is no similar fluctuation repetition within the adjacent 10 seconds, the time point is marked as an "intermittent load node". All marked time points are collected to form an intermittent load node data set, each record point in the data set contains a timestamp, a corresponding load value, and a change direction (increase or decrease). Based on the intermittent load node data, the load values at several time points (±3 second time window) before and after each node are extracted, the stable value L0 before the mutation and the new value L1 after the mutation are recorded, the load mutation amplitude F = |L1-L0| is calculated, and the change direction is recorded. The mutation amplitude data structure is set as a triple: {t, F, dir}, where t is the mutation node time, F is the mutation amplitude (unit kg), and dir is the change direction identifier ("up" or "down"). The effectiveness of the mutation amplitude needs to meet the following conditions: the new load range after the mutation is maintained for at least 5 seconds (with a fluctuation of not more than ±5 kg), otherwise it is judged as short-term noise and not accepted. For the case where the interval between consecutive mutation nodes is less than 10 seconds, only the change with larger amplitude is retained. All valid mutation amplitude samples are numbered and classified to form a complete load mutation amplitude data set. The mutation amplitude data set is statistically classified and abnormality detected. First, calculate the mean μ and standard deviation σ of the overall mutation amplitude, and set the abnormality judgment threshold to μ+1.5σ. All data samples with a mutation amplitude greater than this threshold will be identified as "significant mutation". At the same time, combined with the trolley running stage information, such as acceleration segment, deceleration segment, turning segment, etc., the distribution frequency of the mutation nodes in each stage is counted. If the mutation frequency in a certain time period exceeds three times the normal level (with not more than 1 mutation per minute as the benchmark), the time period will be marked as a "high-frequency mutation segment". Finally, samples that meet one of the following two conditions are selected: the mutation amplitude is greater than the abnormality judgment threshold; or the mutation is in a high-frequency mutation segment, which is defined as "intermittent load mutation feature". These features are uniformly coded, including mutation time, mutation amplitude, change direction, and belonging trajectory position segment, to form the intermittent load mutation feature data used as input for the behavior judgment model.

[0041] Preferably, the personnel behavior feature data of the monkey car passenger in the step S3 is identified, and the behavior pairing and model construction of the monkey car controller behavior feature data and the personnel behavior feature data include: The image preprocessing is performed on the image of the monkey car passenger to generate a definition-optimized image, wherein the image preprocessing includes image brightness enhancement, image geometric transformation and image resolution enhancement; The personnel posture key points are extracted based on the definition-optimized image data to generate personnel posture feature data; The personnel posture feature data is classified to generate personnel behavior feature data; The personnel behavior feature data and the monkey car controller behavior feature data are time-synchronized and behavior-paired to generate behavior-paired data; The controller response judgment model is constructed based on the behavior-paired data to generate a controller response judgment model.

[0042] In the embodiments of the present application, the continuous shooting images of the passengers of the monkey car are collected, the image sampling frequency is set to 25 frames per second, and the image resolution is at least 1920*1080 pixels. In the image preprocessing stage, a high-performance image processing module is used to perform the following operations on the original image: the histogram equalization technology is used to improve the overall brightness and contrast of the image, to ensure clear dark details, and the adjustment parameters are controlled in the range of 20% to 40% to avoid overexposure. The inclination and rotation in the image are corrected by affine transformation, the correction angle is not more than ±5 degrees, the picture is kept stable and true. The bilinear interpolation algorithm is used for super-resolution resampling of the image, the resolution is increased to 1.5 times of the original, and the key part features are ensured to be clear and distinguishable. The open-source multi-human pose recognition algorithm module (such as OpenPose or similar open source library) is used to detect the posture key points of the preprocessed image. The algorithm extracts the joint positions of the human body through a convolutional neural network, the key points include the top of the head, the shoulder, the elbow, the wrist, the hip, the knee and the ankle, etc., a total of 18 standard joint nodes. The algorithm input is a single frame RGB image, and the output is the two-dimensional coordinates and confidence score of the 18 key points. The coordinate data is represented by pixel position, and the confidence threshold is set to 0.6, and the key points below the threshold are automatically removed or interpolated. The personnel posture feature data structure is a set of 18 joint node coordinates for each frame, combined with confidence information storage, used to describe the spatial distribution of the current human posture. The behavior classification adopts a machine learning method based on time series analysis, and the time series of the posture key points of a plurality of continuous frames (recommended 30 frames, about 1.2 seconds) are processed by feature engineering, including joint angle change rate, relative position change between joints, motion speed and acceleration, etc., to form a feature vector. The classification algorithm is based on a trained support vector machine (SVM) or random forest model, and outputs specific behavior category labels such as "normal standing", "walking", "climbing", "falling", etc. The classification decision threshold is set to a confidence probability of 0.8 or more, and low confidence results will be further confirmed by adjacent time frame data. The final personnel behavior feature data includes behavior category, behavior start and end time stamp, behavior duration and corresponding posture feature description. The personnel behavior feature data timestamp and the monkey car controller behavior feature data timestamp are aligned using a unified time reference. The time synchronization error is controlled within ±10 milliseconds, ensuring high-precision pairing. Through the time window matching method, the corresponding controller behavior feature events are selected in the time period when the personnel behavior occurs, to form a behavior event pair. The behavior pairing process adopts a cross-matching algorithm based on time series, which corresponds the personnel behavior category to each index (such as inertia disturbance feature, load mutation feature, etc.) of the controller behavior feature data in the time period. The behavior pairing data structure includes: personnel behavior category, corresponding time period, controller behavior feature index value and its time range, realizing multi-dimensional joint description.Using the paired behavior pairing dataset completed, a multivariate discriminant model is constructed, and a decision tree algorithm or a rule-based expert system is used to determine the controller response. The model input includes the personnel behavior feature category and its time period, the controller behavior feature parameter set (inertial disturbance, load change, blocking delay, etc.), and the output is the controller response state, such as "normal response", "abnormal response", "early warning response", etc. The model parameters are determined by historical data training, and the threshold value such as the abnormal response trigger threshold is set to the controller inertia disturbance index exceeding 0.75 times the standard deviation, the load mutation amplitude exceeding 20 kg, combined with the personnel behavior abnormal category trigger. The controller response judgment result output by the model will be used for subsequent control strategy adjustment and abnormal alarm generation.

[0043] Preferably, the time synchronization and behavior pairing of personnel behavior feature data and controller behavior feature data includes: The personnel behavior feature data and the monkey car controller behavior feature data are respectively divided into time windows to generate multi-level time slice data; The multi-level time slice data is time-aligned to handle the asynchronicity of the two sets of data in the time scale to generate a preliminary time alignment index; The complex action sequence in the personnel behavior feature data and the monkey car controller behavior feature data is decomposed based on a predefined micro-event dictionary to extract extremely fine-grained micro-event sequence data, wherein each micro-event contains action type, duration and intensity features, and a behavior micro-event time sequence is generated; A three-dimensional similarity measurement index is designed in combination with the action type, intensity and time duration features of the micro-event, and a multi-dimensional behavior similarity matrix is self-defined and constructed; According to the preliminary time alignment index and the multi-dimensional behavior similarity matrix, iterative weighted matching is performed to identify the best micro-event corresponding pairing path to generate behavior pairing data.

[0044] In the embodiments of the present application, the time stamp sequences of personnel behavior feature data and monkey car controller behavior feature data are obtained respectively. The time resolution of the two groups of data is usually different, the personnel behavior features are sampled at a second level (frame rate of about 25 frames / s), and the controller data is sampled at a millisecond level (sampling period of 10 milliseconds). Multi-level time window division is performed on the two groups of data respectively, and the time window length is divided into coarse-grained level (for example, 1 second), medium-grained level (200 milliseconds), and fine-grained level (50 milliseconds) according to the level. The statistical values (such as mean value, peak value, and frequency) or event markers of the corresponding behavior features are extracted in each time window to generate the behavior summary corresponding to the time window, forming a multi-level time slice data set. The time window boundary adopts an overlapping sliding window design, and the overlapping ratio is set to 50%, to ensure the continuity and integrity of the time boundary events. Due to the different sampling clocks of the two groups of data, and the network delay and sensor sampling deviation, time synchronization correction is first performed. By cross-correlation analysis on the synchronous event points (such as marked start and end actions or control instruction changes) in the two time sequences, the time offset is calculated to eliminate the system clock difference. For the corrected data, the multi-level time slice data sequence is time-aligned by using the dynamic time warping (DTW) algorithm. DTW allows nonlinear time scaling, which adapts to the time difference between gait and control response. The preliminary time alignment index is output to indicate the correspondence between the personnel behavior time window and the controller behavior time window, providing a time basis for subsequent micro-event matching. A micro-event dictionary is established, including common action types (such as “lift hand”, “bend waist”, “turn around”, etc.) and corresponding controller response events (such as “acceleration peak value”, “motor load mutation”), and each micro-event defines an action type identifier, a duration range (such as 50 milliseconds to 500 milliseconds), and a force feature (such as acceleration amplitude or motor torque size). For the personnel behavior sequence, the posture key point time sequence analysis is used to segment into micro-event units according to the action type and duration threshold. For the controller behavior data, the corresponding micro-event response is marked based on the inertial disturbance peak value, load mutation point, etc., and the control micro-event sequence in the same time period is extracted. The micro-event time sequence is generated, including the action type number, the start and end timestamps, and the force parameter, which structurally represents the fine-grained action and control behavior. The three-dimensional similarity measurement dimensions are defined as follows: action type similarity: based on whether the action type numbers are the same, the Boolean matching is performed, and 1 point is matched, and 0 point is different. Force similarity: the Euclidean distance of the force feature is used for normalized calculation, the smaller the distance, the higher the similarity, the normalized range is 0 to 1, and the distance threshold is set to 0.2. Time duration similarity: the absolute difference of the duration of two micro-events is calculated, normalized, and the time difference is not more than 50 milliseconds, the similarity is 1, and linearly decays to 0. The three-dimensional similarity is combined by weighted summation, and the weights are set to 0.5, 0.3, and 0.2 respectively.The personnel behavior micro-event sequence and the controller behavior micro-event sequence are taken as rows and columns to calculate a similarity matrix, and an element represents a comprehensive similarity score between two micro-events. An improved dynamic programming algorithm is used to search for a path in a time range preliminarily indexed and constrained by time, based on a multi-dimensional similarity matrix, and the goal is to maximize the cumulative sum of similarities on the path, and a path node represents a pairing relationship of a micro-event pair. In the iterative weighted matching process, the time window weight is adjusted according to the matching result of the last round, the search space is dynamically reduced in the poor matching area, and the matching accuracy is improved. Finally, the best micro-event pairing path sequence is output, and structured behavior pairing data is formed, and the data content includes: micro-event number, matching micro-event start and end time, action type matching identifier, force matching score and duration difference. The behavior pairing data is used for subsequent controller response model training and abnormal behavior judgment.

[0045] As an example of the present application, reference is made to Fig. 1, which shows a schematic diagram of a method for abnormal behavior judgment and pre-warning control of a monkey car according to an embodiment of the present application. In this example, the step S4 includes: Figure 3 Step S41: Abnormal feature matching of real-time behavior data of the controller based on the controller response judgment model, to generate abnormal behavior judgment result data; Step S42: Risk level evaluation of the abnormal behavior judgment result data, to generate abnormal behavior risk level data; Step S43: Determination of the corresponding pre-warning control instruction type according to the abnormal behavior risk level data, to generate pre-warning control instruction data, wherein the pre-warning control instruction type includes speed limit, shutdown or alarm; Step S44: Instruction sending to the overhead passenger controller where the monkey car is located through the pre-warning control instruction data, to trigger the corresponding pre-warning control operation.

[0046] ​In the embodiments of the present application, the multi-dimensional behavior data of the overhead passenger controller where the monkey car is located is collected in real time, including but not limited to motor torque, joint angle, load change and inertia sensor data. The collected real-time data is input into the trained controller response judgment model according to the preset sampling frequency (such as every 10 milliseconds). The model is based on machine learning classification algorithms (such as support vector machine, random forest or lightweight neural network) and has learned the feature distribution of normal and abnormal control behavior in advance. For the input real-time data, the model performs multi-feature fusion and matching operation, and judges whether the current behavior deviates from the normal range through the abnormality score function in the feature space. The abnormal threshold is set to be considered as abnormal behavior when the abnormality score exceeds 0.7 (full score 1). The output abnormal behavior judgment result data includes abnormal timestamp, abnormal feature type (such as overload, stall, shock, etc.), abnormality score and associated parameter value, forming a structured abnormality judgment log. Based on the abnormal behavior judgment result, a hierarchical risk assessment algorithm is used to quantify the severity of the abnormal behavior. The risk level is divided into three levels: first-level risk (high risk): abnormality score ≥ 0.9, and abnormal features related to safety-related indicators, such as motor overload exceeding limit, current abnormal fluctuation amplitude exceeding 20%, etc. Second-level risk (medium risk): abnormality score between 0.7 and 0.9, abnormal features are functional fluctuations, such as short-term stall or slight vibration. Third-level risk (low risk): abnormality score is less than 0.7, but there are non-critical indicators deviating. When assessing the risk level, the duration of the abnormality is considered, and the continuous abnormality exceeding 500 milliseconds will increase the level. Combined with the type and duration of the abnormality, the risk level data of the abnormal behavior is generated, including risk level number, evaluation basis and recommended processing level. According to the risk level data, the pre-defined warning control instruction strategy library is matched. The warning control instruction type is divided into three types: speed limiting instruction: suitable for second-level risk, control the monkey car speed to reduce by 30% to 50%, the specific speed limiting ratio is determined according to the actual risk level grading. Shutdown instruction: suitable for first-level risk, immediately stop the monkey car running to prevent accidents. Alarm instruction: suitable for third-level risk, trigger remote or on-site alarm notification, without affecting the current operation. The instruction generation module combines the current controller state and safety specifications to form structured warning control instruction data, including instruction type, effective time, target device ID and related execution parameters. The warning control instruction is sent to the overhead passenger controller where the monkey car is located through a real-time communication link (such as industrial Ethernet or wireless dedicated channel), and the communication protocol uses a secure authentication transmission layer protocol to ensure data integrity and timeliness. After receiving the instruction, the controller analyzes the instruction type and parameters, and immediately executes the corresponding warning control operation, including speed limiting control signal issuance, mechanical emergency stop triggering or alarm sound and light device activation. The feedback signals in the execution process are monitored in real time to confirm whether the instruction is executed successfully, and the abnormal feedback information is recorded for subsequent fault diagnosis. After the warning control operation is completed, the system automatically enters the monitoring state, waiting for subsequent instructions to ensure the safe operation of the monkey car.

[0047] Thus, the embodiments should be considered in all respects as illustrative and not restrictive, the scope of the application being indicated by the appended claims rather than by the foregoing description, and all changes which come within the meaning and range of equivalency of the claims are intended to be embraced therein.

[0048] The foregoing is considered as illustrative only of the principles of the application. Numerous modifications and changes will readily occur to those skilled in the art, and the generic principles defined herein can be applied to other embodiments without departing from the spirit or scope of the application. Therefore, the application is not intended to be limited to the embodiments shown herein but is to be accorded the widest scope consistent with the principles and novel features disclosed herein.

Claims

1. A data processing method for an aerial passenger controller, characterized in that, Includes the following steps: Step S1: Obtain the location information data of the monkey bus and the controller data collected by the overhead passenger controller where the monkey bus is located; confirm the running trajectory of the monkey bus based on the location information data, and perform three-dimensional modeling on the controller data based on the running trajectory of the monkey bus to generate three-dimensional modeling data of the monkey bus; Step S2: Analyze the inertial disturbance characteristics, inertial stagnation delay characteristics, and discontinuous load mutation characteristics of the 3D modeling data of the monkey car, and perform data feature distribution fusion to generate a monkey car controller behavior feature dataset. Step S3: Acquire images of the occupants of the monkey bus; identify the behavioral characteristics of the occupants in the images, and perform behavioral pairing and model building between the behavioral characteristics of the monkey bus controller and the behavioral characteristics of the occupants to generate a controller response judgment model; Step S4: Use the controller response judgment model to identify abnormal behavior of the controller. When abnormal behavior is detected, issue speed limit, stop or alarm commands to the overhead passenger controller where the monkey bus is located to execute the corresponding early warning control operation.

2. The data processing method for an aerial passenger controller according to claim 1, characterized in that, Step S1 includes the following steps: Step S11: Obtain the location information data of the monkey bus and the controller data collected by the overhead passenger controller where the monkey bus is located; Step S12: Reconstruct the coordinate trajectory of the monkey vehicle's location information data to generate the monkey vehicle's three-dimensional running trajectory data; Step S13: Extract the controller timestamp from the controller's collected data, and combine it with the monkey car's 3D running trajectory data to perform trajectory point mapping and fusion, generating a trajectory-bound controller dataset; Step S14: Perform spatial data interpolation and parameter normalization based on the trajectory binding controller dataset to generate trajectory control parameter distribution data; perform spatial mesh modeling based on the trajectory control parameter distribution data to generate 3D modeling data for the monkey car.

3. The data processing method for an aerial passenger controller according to claim 2, characterized in that, Step S14, which involves spatial grid modeling based on trajectory control parameter distribution data, includes: Based on the trajectory control parameter distribution data, spatial discretization is performed to divide it into several three-dimensional grid units to generate initial spatial grid data. Boundary conditions are identified in the initial spatial grid data, trajectory-related grid cells are labeled, and trajectory-constrained grid data is generated. The spatial attribute parameters of each grid cell are calculated based on the trajectory-constrained grid data, including position coordinates and motion state indicators, to generate grid attribute parameter data; By using grid attribute parameter data to establish topological connections between grid cells, a complete spatial grid model is constructed, generating 3D modeling data for the monkey car.

4. The data processing method for an aerial passenger controller according to claim 1, characterized in that, Step S2 analyzes the inertial perturbation characteristics of the 3D modeling data of the monkey cart, including: Acceleration and angular velocity signals are extracted from the 3D modeling data of the monkey vehicle to generate inertial motion signal data; The inertial motion signal data is subjected to time-domain filtering to remove high-frequency noise and generate filtered inertial signal data. The rate of change of acceleration and the rate of change of angular velocity are calculated using the filtered inertial signal data to obtain the amplitude data of inertial disturbance. Statistical analysis was performed on the amplitude data of inertial disturbances to extract characteristic inertial disturbance indicators and obtain the characteristics of inertial disturbances.

5. The data processing method for an aerial passenger controller according to claim 1, characterized in that, Step S2 analyzes the inertial lag delay characteristics of the 3D modeling data of the monkey cart, including: Acceleration and angular velocity signals are extracted from the 3D modeling data of the monkey vehicle to generate inertial motion signal data; Based on inertial motion signal data, the time delay of the inertial lag response is calculated to obtain inertial lag delay data; Time-domain and frequency-domain analyses were performed on the inertial hysteresis delay data to extract delay characteristic parameters and generate inertial hysteresis characteristic parameter data. By combining the inertial stagnation characteristic parameter data, the stagnation effect on the motion state of the monkey car is evaluated, and the inertial stagnation delay characteristics are obtained.

6. The data processing method for an aerial passenger controller according to claim 1, characterized in that, Step S2 analyzes the discontinuous load mutation characteristics of the monkey cart 3D modeling data, including: Load change time series is extracted from the 3D modeling data of the monkey cart to generate load time series data; Perform segmented statistical analysis on load time-series data to identify discontinuous nodes of load changes and generate discontinuous load node data. Calculate the load fluctuation amplitude at the discontinuous load node and generate load fluctuation amplitude data; Abnormal fluctuations are detected based on load mutation amplitude data, and significant discontinuous load mutation characteristics are screened out to obtain discontinuous load mutation characteristics.

7. The data processing method for an aerial passenger controller according to claim 1, characterized in that, Step S3 involves identifying the behavioral characteristics of the passengers in the images captured by the monkey bus, and then matching the behavioral characteristics of the monkey bus controller with the behavioral characteristics of the passengers to build a model. Image preprocessing is performed on images of passengers in the monkey bus to generate images with improved clarity. Image preprocessing includes image brightness enhancement, image geometric transformation, and image resolution enhancement. Based on the image data with improved clarity, key points of human posture are extracted to generate human posture feature data. Perform behavioral classification on personnel posture feature data to generate personnel behavioral feature data; The human behavior characteristic data is synchronized with the monkey bus controller behavior characteristic data in time and the behavior is matched to generate behavior matching data. A controller response judgment model is constructed based on behavior pairing data, and the controller response judgment model is generated.

8. The data processing method for an aerial passenger controller according to claim 7, characterized in that, Synchronizing and matching personnel behavior data with controller behavior data in time includes: The behavioral characteristic data of personnel and the behavioral characteristic data of the monkey bus controller are divided into time windows to generate multi-level time slice data. Time alignment is performed on multi-level time slice data to handle the asynchrony of the two sets of data on the time scale and generate a preliminary time alignment index. Complex action sequences from personnel behavior feature data and monkey bus controller behavior feature data are decomposed based on a predefined micro-event dictionary to extract extremely fine-grained micro-event sequence data. Each micro-event contains action type, duration, and intensity features, generating a behavioral micro-event time series. By combining the action type, intensity, and duration characteristics of micro-events, a three-dimensional similarity metric is designed, and a custom multi-dimensional behavior similarity matrix is ​​constructed. Iterative weighted matching is performed based on the initial time alignment index and the multidimensional behavior similarity matrix to identify the best micro-event corresponding pairing path and generate behavior pairing data.

9. The data processing method for an aerial passenger controller according to claim 1, characterized in that, Step S4 includes the following steps: Step S41: Based on the controller response judgment model, perform abnormal feature matching on the real-time behavior data of the controller to generate abnormal behavior discrimination result data; Step S42: Assess the risk level of the abnormal behavior identification results data and generate abnormal behavior risk level data; Step S43: Based on the abnormal behavior risk level data, determine the corresponding early warning control instruction type and generate early warning control instruction data, wherein the early warning control instruction type includes speed limit, shutdown or alarm; Step S44: Send a warning control command to the overhead passenger controller where the monkey bus is located via the warning control command data to trigger the corresponding warning control operation.

10. A data processing system for an aerial passenger controller, characterized in that, For performing the data processing method for an aerial passenger controller as described in claim 1, the data processing system for the aerial passenger controller includes: The 3D modeling module is used to acquire the location information data of the monkey bus and the controller data collected by the overhead passenger controller where the monkey bus is located; based on the location information data of the monkey bus, the running trajectory of the monkey bus is confirmed, and 3D modeling is performed on the controller data based on the running trajectory of the monkey bus to generate 3D modeling data of the monkey bus; The control behavior feature module is used to analyze the inertial disturbance features, inertial stagnation delay features, and discontinuous load mutation features of the 3D modeling data of the monkey car, and to perform data feature distribution fusion to generate a monkey car controller behavior feature dataset. The user behavior feature module is used to acquire images of the occupants of the monkey bus; identify the human behavior feature data in the images of the occupants of the monkey bus; and perform behavior matching and model building between the monkey bus controller behavior feature data and the human behavior feature data to generate a controller response judgment model. The early warning control module is used to identify abnormal behavior of the controller by using the controller response judgment model. When abnormal behavior is detected, it sends speed limit, shutdown or alarm commands to the overhead passenger controller where the monkey bus is located to execute the corresponding early warning control operation.

Images