Railway container FTR lock operation and unbalance load and weight monitoring and early warning system and method

Abstract

This invention discloses a system and method for monitoring and early warning of FTR lock operation and off-center loading / unloading on railway containers. The system includes a pre-monitoring module, a critical pre-screening module, a multi-level interception module, and an active control module. The pre-monitoring module generates pre-monitoring information based on historical operation data, cross-domain operation data, and environmental risk data before the container arrives. The critical pre-screening module performs stability prediction when the spreader locking pin is inserted into the FTR lock hole but not fully locked. The multi-level interception module progressively detects the FTR lock operation status and container attitude along the descent path of the spreader on the railway container loading and unloading equipment. The active control module switches to active correction mode and applies a target correction torque when a center of gravity deviation is detected. This invention comprehensively improves the safety and efficiency of container operations by constructing a time-progressive, in-depth defense system consisting of predictive data loading, critical window pre-screening, multi-level progressive interception, and active correction control.

Description

Technical Field

[0001] This invention relates to the field of safety technology for railway container loading and unloading operations, specifically to a system and method for monitoring the operational status of container FTR locks and providing early warning of container overloading / overweight issues, applicable to the "one car, two containers" transportation conditions of railway flatcars. Background Technology

[0002] Railway container transport has become a core carrier of inland and international logistics. Railway freight yards and container terminals, as key nodes in railway container transport, directly impact the operation of the entire logistics chain and train safety through their loading and unloading efficiency and safety. In railway container loading and unloading operations, FTR locks (Fully Twistlock Ready) are widely used as standardized, fully automatic twist-locking devices for connecting containers to loading and unloading equipment (such as gantry cranes and reach stackers) and to railway flatcars. Reliable locking of FTR locks is a prerequisite for ensuring the stability of containers during loading, unloading, and train operation.

[0003] In recent years, to meet the growing demand for China-Europe freight train services, new types of dedicated railway container flatcars have emerged. Taking the first scheduled China-Europe freight train using the BX70B type container flatcar, launched on April 8, 2025, by the Shaliang Logistics Park of China Railway Hohhot Bureau Group Co., Ltd., as an example, the BX70B flatcar is 26.366 meters long, has a load capacity of 68 tons, a tare weight of 25.8 tons, a design speed of 120 km / h, and can simultaneously carry two 40-foot containers, doubling the cargo capacity of ordinary flatcars and achieving a highly efficient "one car, two containers" transportation model.

[0004] However, while improving transportation efficiency, the "one car, two containers" transport mode also brings new challenges to monitoring off-center loading and weight distribution. When two containers are loaded onto the same flatcar, the following technical issues arise: the center-of-gravity shifts of the two containers can couple, resulting in a combined shift of the flatcar's center of gravity. When the weight difference between the two containers is too large or their centers of gravity shift in the same direction, the combined center of gravity may exceed the safe range, directly threatening train safety. Single-container off-center loading (lateral shift of the center of gravity) may have a cumulative effect in the "one car, two containers" mode. Even if the single-container off-center loading is within the allowable range, the combined off-center loading of both containers may still exceed the limit. Railway freight regulations have strict limits on container off-center loading and weight distribution. In the "one car, two containers" mode, it is necessary not only to monitor the weight of a single container but also the weight difference between the two containers, the latter directly affecting the wheel weight distribution of the flatcar bogie. When the types of goods, binding methods, and center-of-gravity distributions of the two containers are different, uncoordinated dynamic responses may occur during lifting or lowering, leading to problems such as container swaying and uneven force on the locking devices.

[0005] Existing railway container handling monitoring systems suffer from the following problems: Firstly, existing systems often only detect off-center loads after the container has been fully lifted and unloaded or placed back on the platform. At this point, it's difficult to effectively correct the anomaly, forcing a shutdown of operations and missing the optimal window for risk management. This also severely impacts train dwell time at stations and track capacity. Secondly, existing monitoring methods are fragmented and independent, resulting in significant data silos. A complete protection chain from prediction, pre-screening, interception to proactive control is not formed, and there is a lack of coordination between each link. Furthermore, the system fails to consider the need for inter-container collaborative monitoring under the "one car, two containers" scenario. Thirdly, when off-center loads are detected, existing systems often adopt passive adaptation or simply prohibit lifting, failing to fully utilize the proactive control capabilities of loading and unloading equipment. In the "one car, two containers" scenario, the current... Some systems lack control strategies for coordinated correction of two containers, failing to simultaneously meet the dual constraints of single-container balance and overall wheel load distribution of the flatcar. Existing systems are mostly designed for "one car, one container," failing to consider the coupling effect of the center of gravity between containers when two containers are loaded on the same flatcar, the risk of overlapping off-center loads of the two containers, and the impact of the overall center of gravity shift of the flatcar on the wheel load distribution of the bogie. When the weight difference between the two containers is greater than 3 tons or the combined off-center load is greater than 100 mm, it will directly threaten train operation safety, and existing systems cannot effectively identify and handle such risks. Existing systems generally use fixed threshold judgments, failing to consider the impact of railway-specific environmental factors such as track stiffness conditions, wind zone environment, total train weight, and "one car, two containers" working conditions on safety thresholds, resulting in decreased early warning accuracy under complex working conditions.

[0006] In summary, existing technologies have significant shortcomings in terms of early warning timing, defense-in-depth architecture, proactive control strategies, adaptability to "one car, two boxes" working conditions, and environmental adaptability, making it difficult to meet the intelligent, efficient, and safe loading and unloading operations required by modern railway freight yards. Summary of the Invention

[0007] The purpose of this invention is to provide a railway container FTR lock operation and off-center load monitoring and early warning system and method, which realizes predictive monitoring, critical state pre-screening, multi-level progressive interception and dual-container collaborative active correction control of the entire railway container loading and unloading process under the "one car, two containers" transportation condition of railway flatcars, so as to comprehensively improve the safety and efficiency of railway container loading and unloading operations.

[0008] To solve the above-mentioned technical problems, the technical solution of the present invention is as follows:

[0009] A railway container FTR lock operation and off-center load monitoring and early warning system includes a pre-monitoring module, a critical pre-screening module, a multi-level interception module and an active control module;

[0010] The pre-monitoring module is used to generate pre-monitoring information based on historical operation data, cross-domain operation data, and environmental risk data before the containers arrive at the railway freight yard operation area, and push the pre-monitoring information to the downstream module; the pre-monitoring module receives double container formation information from the train formation information system, generates a double container joint pre-loading operation file, and performs a container center of gravity coupling risk integral calculation on the two containers loaded on the same flatcar. When the container center of gravity coupling risk integral exceeds the threshold, a double container collaborative enhanced monitoring marker is triggered.

[0011] The critical pre-screening module is connected to the pre-monitoring module and is used to perform stability prediction on the two containers based on the pre-monitoring information when the locking pin of the spreader of the railway container loading and unloading equipment is inserted into the FTR lock hole but not fully locked, and output the prediction result. The multi-level interception module is connected to the critical pre-screening module and is used to progressively detect the FTR lock operation status and container attitude of the two containers along the descent path of the spreader of the railway container loading and unloading equipment when the prediction result shows that the operation is stable. If any detection point is abnormal, the operation path is locked.

[0012] The active control module is connected to the critical pre-screening module and the multi-level interception module respectively. It is used to receive the prediction results or the off-seat / off-ground posture data of the progressive detection, calculate the single-container center of gravity offset of the two containers respectively and synthesize the overall center of gravity offset of the flatcar. When the overall center of gravity offset of the flatcar exceeds the flatcar center of gravity threshold or the weight difference between the two containers exceeds the weight difference threshold, the dual-container collaborative correction mode is triggered, and the target correction torque is applied so that the two containers reach the critical balance state at the moment of lifting or sitting.

[0013] The aforementioned pre-monitoring module, critical pre-screening module, multi-level interception module, and active control module form a time-series progressive defense system from prediction, pre-screening, interception to active control.

[0014] Furthermore, the pre-monitoring module includes a pre-loaded data cache pool, a cross-domain attitude database, and a risk prediction unit; the pre-loaded data cache pool receives double-box marshalling information from the railway freight station management system, historical center of gravity offset records from the historical database, and lock wear parameters from the FTR lock status monitoring system, and generates a double-box joint pre-loading operation file;

[0015] The cross-domain attitude database receives real-time operation data broadcast from monitoring units in other sections of adjacent tracks via a data bus listening mode. This data includes container type identification, lifting attitude angular velocity sequence, and FTR lock engagement anomaly markers. Local sensitive hash similarity indexing is performed on the real-time operation data to establish similar container type attitude data with container type identification as the primary key and attitude angular velocity distribution as the feature value. This data is then sent to the pre-loaded data cache pool to be associated with the corresponding pre-loaded operation file. When the container type identification of the target container in this loading / unloading line operation matches the similar container type attitude data, the corresponding attitude angular velocity distribution is extracted as a pre-rehearsal reference benchmark for this container operation.

[0016] The risk prediction unit receives train formation information from the railway freight station management system, historical train load imbalance data from the railway bureau's historical operation data center, and wind zone weather forecast information from the railway meteorological station interface. It performs risk score calculation for each container in the train formation information, outputs a load imbalance risk heat map, and pushes the high-risk container identifier with a risk score exceeding the threshold in the load imbalance risk heat map to the preloaded data cache pool to trigger enhanced monitoring.

[0017] Furthermore, the critical pre-screening module independently applies tentative micro-disturbances to the two containers, collects the acceleration response sequence of the two containers, and extracts the joint natural frequency and damping ratio of the two containers through inter-container coupling mode identification. When the offset of the coupling frequency of the two containers exceeds the threshold or the phase difference between the containers exceeds the allowable range, it is determined that the loading between the containers is not coordinated, outputs the stability failure prediction result, and terminates the current operation process. The modal parameter extraction of the tentative micro-disturbance also includes a track baseline mode subtraction step: before identifying the modal parameters of the container-spreader system, the track-vehicle coupled vibration baseline signal under the empty state of the railway flatcar is first collected, and the interference of rail irregularities and track bed elastic deformation on the acceleration response sequence is eliminated by frequency domain subtraction or adaptive filtering.

[0018] Furthermore, the calculation of the target correction torque also incorporates the stiffness compensation coefficient of the railway flatcar suspension system: when the container is placed on the railway flatcar, the differential output of the hydraulic cylinder is corrected based on the real-time stiffness feedback of the air spring / steel spring of the vehicle bogie to compensate for the attitude error caused by the elastic deformation of the car body frame.

[0019] Furthermore, the multi-level interception module is a multi-level interception detection unit deployed along the descent path of the spreader of the railway container loading and unloading equipment. It includes a first-level mechanical alignment detection point, a second-level locking pin engagement detection point, and a third-level locking pin engagement detection point. The first-level mechanical alignment detection point receives the mechanical alignment deviation value of the dual-box FTR lock from the spreader's visual alignment system. When the mechanical alignment deviation value exceeds the allowable range, a first-level abnormal signal is output. If no abnormality is detected, the second-level locking pin engagement detection point is activated. The second-level locking pin engagement detection point receives the engagement depth value and engagement uniformity of the dual-box FTR lock from the locking pin pressure sensor. The index outputs a second-level abnormal signal when the engagement depth value is lower than the safe depth or the engagement uniformity index exceeds the non-uniformity threshold; if no abnormality is determined, the third-level off-seat / off-ground attitude detection point is activated; the third-level off-seat / off-ground attitude detection point receives the angular acceleration value of the two containers leaving the seat or leaving the ground from the spreader tilt angle sensor, and outputs a third-level abnormal signal when the angular acceleration value exceeds the dynamic imbalance threshold; the multi-level interception detection unit generates a path locking command when it receives any level abnormal signal, prohibiting the spreader of the railway container loading and unloading equipment from continuing to operate until manual confirmation and re-execution of this level and all subsequent detections.

[0020] Furthermore, it also includes an internal and external sensor fusion unit; the internal and external sensor fusion unit includes: a passive RFID pressure sensing tag deployed on the bottom surface inside the container, which is passively activated by the electromagnetic field of the RFID reader at the end of the spreader during the lifting process of the railway container loading and unloading equipment, collects the pressure field data of the cargo distribution inside the container and transmits it through backscatter modulation; a stress-strain sensor embedded inside the FTR lock pin, which is self-powered through the piezoelectric effect of the piezoelectric material inside the lock pin, and collects the three-dimensional force state data of the lock pin; a coupled force model generator, which receives the pressure field data and the three-dimensional force state data, performs data spatiotemporal alignment processing with the container number as the associated primary key, the system unified clock as the reference, and ±100ms as the time window matching rule, and establishes a container-lock coupled force model to identify the mismatch pattern between the internal cargo off-center loading and the external lock force.

[0021] Furthermore, it also includes an environment-adaptive threshold unit; the environment-adaptive threshold unit receives track stiffness coefficients from the track monitoring system of the railway engineering section, wind speed and direction data from meteorological stations along the railway line, and total weight and axle load distribution data of the train formation information system, queries the built-in dynamic threshold correction matrix, and outputs the corrected off-center load warning threshold and off-center weight warning threshold; the dimensions of the dynamic threshold correction matrix include: track stiffness level, wind zone level range, and total weight range of the train formation.

[0022] Furthermore, the system is applicable to BX70B type container flatcars: the risk integral R between container centers of gravity coupling couple=|W1-W2| / 3+|X1+X2| / 0.2, where W1 and W2 are the weights of the two boxes, and X1 and X2 are the lateral offsets of the center of gravity of a single box; the threshold for the coupling frequency offset of the two boxes is 0.20, and the allowable range for the phase difference between the boxes is 45 degrees; the threshold for the center of gravity of the flatcar is 100 mm, and the threshold for the weight difference is 3 tons; the second-level locking pin engagement detection point of the multi-level interception module receives 8 sets of FTR lock engagement depth values ​​and engagement uniformity index.

[0023] A method for monitoring and early warning of FTR lock operation and eccentric loading / weight imbalance in railway containers, based on the aforementioned monitoring and early warning system, includes the following steps:

[0024] The pre-monitoring steps involve generating pre-monitoring information based on historical operation data, cross-domain operation data, and environmental risk data before containers arrive at the railway freight yard operation area; receiving double container marshalling information and generating a double container joint pre-loading operation file; performing inter-container center of gravity coupling risk integral calculation, and triggering a double container collaborative enhanced monitoring marker when the threshold is exceeded.

[0025] The critical pre-screening step involves making stability predictions for two containers in a critical state where the locking pin of the spreader of the railway container loading and unloading equipment is inserted into the FTR lock hole but not fully locked.

[0026] The multi-level interception process involves progressively detecting the FTR lock operation status and container attitude of the two containers along the spreader descent path when the prediction result indicates stability.

[0027] The active control step receives the prediction result or the off-seat / off-ground posture data of the progressive detection, calculates the single-container center of gravity offset of the two containers and synthesizes the overall center of gravity offset of the flatcar. When the overall center of gravity offset of the flatcar exceeds the flatcar center of gravity threshold or the weight difference between the two containers exceeds the weight difference threshold, the dual-container collaborative correction mode is triggered, and the target correction torque is applied so that the two containers reach the critical balance state at the moment of lifting or sitting.

[0028] The aforementioned pre-monitoring steps, critical pre-screening steps, multi-level interception steps, and active control steps form a time-series progressive defense-in-depth process from prediction, pre-screening, interception to active control.

[0029] Furthermore, the critical pre-screening step and the multi-level interception step complement each other in timing: the critical pre-screening step completes risk pre-screening during the first critical window period when the spreader contacts the container but is not completely locked, and the multi-level interception step completes multi-level interception during the second critical window period when the spreader descends and lifts off the seat / lifts off the ground. The two steps cover different risk exposure periods in the operation process, and the termination decision of the critical pre-screening step takes precedence over the execution of the multi-level interception step.

[0030] The present invention achieves the following beneficial effects through the above technical solution:

[0031] This invention generates joint pre-loading operation files for two containers before they arrive at the railway freight yard through a pre-monitoring module, transforming the traditional real-time query mode into a pre-calculation mode. This mechanism avoids database access congestion during peak operation periods, allowing the system to immediately call pre-stored data to perform joint matching of two containers when they enter the operation area, significantly improving response speed. Through risk integral calculation of the coupling of center of gravity between containers, the system identifies the weight difference between the two containers and the risk of combined center of gravity shift before the train enters the station, transforming post-event detection into pre-event prediction. This enables the system to deploy enhanced monitoring resources in advance for high-risk two-container combinations, achieving optimized allocation of monitoring resources.

[0032] By applying tentative micro-disturbances independently to two containers during the critical window period when the locking pins are inserted but not fully locked using a critical pre-screening module, the stability of lifting or seating is predicted using inter-container coupling mode recognition. This advances the critical time window compared to traditional post-lift detection, allowing sufficient time for risk management. The track baseline mode subtraction mechanism effectively eliminates interference from rail irregularities and track bed elastic deformation on modal parameter recognition, improving the accuracy of predictions under complex railway track conditions. When the coupling frequency offset between the two containers exceeds the threshold or the phase difference between the containers exceeds the allowable range, the system can identify the risk of incoordination in loading between the containers in advance, avoiding operations with potential hazards.

[0033] Through a three-level progressive detection design of multi-level interception and detection units, deep defense is achieved in the FTR lock operation process under the "one vehicle, two boxes" working condition. The first level mechanical alignment detection solves the problem of spatial position matching between the two boxes; the second level lock pin engagement detection solves the problem of mechanical connection reliability of the eight sets of FTR locks in the two boxes; and the third level off-seat / off-ground posture detection solves the problem of dynamic stability of the two boxes. The three levels of detection cover different risk types and are independent of each other. The progressive judgment logic avoids redundant detection, and the path locking command ensures the safety of operation under abnormal conditions.

[0034] By using the active control module to calculate the combined center of gravity of the two containers and constrain the overall wheel load distribution of the flatcar, a shift from passive adaptation to active intervention is achieved. The system calculates the center of gravity offset of each of the two containers separately, synthesizes the overall center of gravity offset of the flatcar, and triggers a dual-container collaborative correction mode when the overall center of gravity offset or the weight difference between the two containers exceeds the limit. The hydraulic leveling mechanism of the loading and unloading equipment actively applies a correction torque, ensuring that the two containers reach a critical equilibrium state at the moment of lifting or placement, significantly improving operational efficiency compared to the traditional method of prohibiting lifting and then manually handling the situation. The target correction torque must simultaneously satisfy the single-container equilibrium condition and the overall wheel load distribution condition of the flatcar, solving the problem of balancing the torque distribution between containers in the "one truck, two containers" working condition. The flatcar suspension stiffness compensation mechanism ensures that the elastic deformation of the car body underframe during placement does not introduce additional attitude errors.

[0035] By integrating passive RFID pressure sensing tags and stress-strain sensors into an internal and external sensing fusion unit, synchronous monitoring of the container's internal condition and the force exerted on the external locks is achieved. This mechanism overcomes the limitations of traditional external monitoring perspectives. By constructing a container-lock coupled force model, it can identify mismatch patterns between internal cargo off-center loading and external lock force, improving the accuracy of off-center loading identification. The passive design and piezoelectric self-powered design avoid the battery maintenance issues of the sensors.

[0036] By using the dynamic threshold correction matrix of the environmental adaptive threshold unit, the early warning threshold is adjusted adaptively to the environment. This mechanism takes into account multiple railway-specific environmental factors such as track stiffness, wind zone environment, and total weight of train formation. It automatically adjusts the threshold sensitivity under complex working conditions and is more adaptable to changes in the actual railway loading and unloading operation environment than a fixed threshold, thus reducing false alarms and missed alarms. Attached Figure Description

[0037] Figure 1 This is a schematic diagram of the monitoring and early warning system architecture of the present invention;

[0038] Figure 2 This is a schematic diagram of the deployment of three-level detection points in the multi-level interception and detection unit of the present invention;

[0039] Figure 3 This is a flowchart of the working condition execution method of the present invention. Detailed Implementation

[0040] To facilitate understanding of the present invention, the invention will be further described in detail below with reference to the accompanying drawings and specific embodiments. Those skilled in the art should understand that the embodiments described are merely illustrative of the invention and should not be considered as specific limitations thereof.

[0041] In the following description of the embodiments, specific details such as particular system architectures and techniques are set forth for illustrative purposes and not for limitation, in order to provide a thorough understanding of the embodiments of this application. However, those skilled in the art will understand that this application may also be implemented in other embodiments without these specific details. In other instances, detailed descriptions of well-known systems, apparatuses, circuits, and methods are omitted so as not to obscure the description of this application with unnecessary detail.

[0042] It should be understood that, when used in this specification and the appended claims, the term "comprising" indicates the presence of a described feature, integral, step, operation, element, and / or component, but does not exclude the presence or addition of one or more other features, integrals, steps, operations, elements, components, and / or collections thereof. It should also be understood that, as used in this specification and the appended claims, the term "and / or" refers to any combination of one or more of the associated listed items and all possible combinations, and includes such combinations.

[0043] As used in this application specification and the appended claims, the term "if" may be interpreted, depending on the context, as "when," "once," "in response to determination," or "in response to detection." Similarly, the phrase "if determined" or "if detected [the described condition or event]" may be interpreted, depending on the context, as "once determined," "in response to determination," "once detected [the described condition or event]," or "in response to detection [the described condition or event]."

[0044] Furthermore, in the description of this application and the appended claims, the terms "first," "second," "third," etc., are used only to distinguish descriptions and should not be construed as indicating or implying relative importance. References to "one embodiment" or "some embodiments" in this application mean that one or more embodiments of this application include a specific feature, structure, or characteristic described in connection with that embodiment. Therefore, the phrases "in one embodiment," "in some embodiments," "in other embodiments," "in still other embodiments," etc., appearing in different parts of this specification do not necessarily refer to the same embodiment, but rather mean "one or more, but not all, embodiments," unless otherwise specifically emphasized. The terms "comprising," "including," "having," and variations thereof all mean "including but not limited to," unless otherwise specifically emphasized.

[0045] Example 1:

[0046] like Figure 1 As shown, the container FTR lock operation and off-center load monitoring and early warning system for the "one car, two containers" transportation condition of railway flatcars in this embodiment includes a pre-monitoring module, a critical pre-screening module, a multi-level interception module and an active control module, as well as an optional internal and external sensor fusion unit and an environmental adaptive threshold unit.

[0047] This embodiment applies to the "one car, two containers" transportation scenario of general railway flatcars, that is, two containers are loaded on the same railway flatcar at the same time (which can be two 20-foot containers, two 40-foot containers, or a combination of one 20-foot and one 40-foot container). The system is designed for the collaborative monitoring of two containers.

[0048] (i) The pre-monitoring module includes a pre-loaded data cache pool, a cross-domain attitude database, and a risk prediction unit. The pre-loaded data cache pool is deployed in the memory area of ​​the industrial computer (IPC) in the railway freight yard control building or station dispatching room, and establishes data connection with the railway freight station management system (TMIS), the railway historical operation database, and the FTR lock status monitoring system through an Ethernet interface.

[0049] The preloaded data cache pool receives three types of data: 1) Double-box formation information: Train formation information / freight bill data messages from the railway freight station management system, including flatcar number, double-box number, double-box type (20 feet / 40 feet / 45 feet), double-box cargo code, double-box rated load, double-box tare weight, double-box size parameters, double-box weight difference warning threshold, and combined center of gravity offset threshold; 2) Historical center of gravity offset records: From the railway historical operation database, including the calculated coordinates of the center of gravity of the double-box number in the past operations of the railway freight yard, off-center load alarm records, and statistics on the group center of gravity distribution of similar double-box combinations; 3) Lock wear parameters: From the maintenance files of the FTR lock status monitoring system, including the cumulative number of times the FTR lock pin of the double-box has been engaged, the time of the most recent maintenance, and the estimated wear amount.

[0050] The preloaded data cache pool performs pre-association processing on these three types of data to generate a dual-bin joint preload job file:

[0051] Data cleaning: The format of the double-box grouping information is validated, and records with incorrect box number format (not conforming to GB / T38946-2020 standard) or missing key fields are removed; historical center of gravity offset records are processed by time decay, and records older than 12 months are assigned lower weights.

[0052] Double-container association analysis: Historical center of gravity offset records are clustered according to double-container combination type, single-container type, cargo type, and weight class. Clustering dimensions include: double-container combination type (same type double 20-foot / same type double 40-foot / different type combination), single-container cargo type (dry / refrigerated / open top / frame), and single-container weight class (light load <10t / medium load 10-20t / heavy load >20t).

[0053] Calculation of inter-container center-of-gravity coupling risk integral: For each pair of containers, calculate the inter-container center-of-gravity coupling risk integral R. couple :

[0054] R couple =α·|W1-W2| / ΔW max +β·|X1+X2| / ΔX max +γ·ρ cargo ;

[0055] In the formula: W1 and W2 are the weights of the two containers (tons), X1 and X2 are the lateral offsets of the center of gravity of a single container (meters), obtained from historical data statistics or preliminary estimates, ΔW max The warning threshold for the weight difference between the two containers (general value 5 tons, BX70B model 3 tons), ΔX max The composite center of gravity offset threshold (general value 150 mm, BX70B model 100 mm), ρ cargoα is a coupling term for the anomaly of the weight-to-volume ratio of double-box cargo, and α, β, and γ are weighting coefficients, which are set to 0.4, 0.4, and 0.2 in this embodiment.

[0056] When R couple When the value is ≥1.0, the dual-box collaborative enhanced monitoring flag is triggered (the flag is_dual_enhanced=true is set) and pushed to the downstream module.

[0057] File generation: The double-car train grouping information, double-car train center of gravity feature model, lock wear parameters and confidence level markers are packaged into a double-car train joint pre-loading operation file, which resides in the IPC memory in the form of a memory-mapped file and is set to a lifespan of 24 hours after the train leaves the station.

[0058] The formula for calculating the preloading timing T is: T = T TMIS -Δt load -Δt margin T TMIS The estimated arrival and departure times of trains reported by the Railway Station Freight Management System (TMIS), Δt load The preloading processing time (dynamically estimated based on the data volume, typically 5-15 minutes), Δt margin For safety margin (fixed at 10 minutes).

[0059] The cross-domain attitude database is deployed on distributed storage nodes in the railway bureau's data center and connects to the control buildings or station dispatching rooms of each railway freight yard via a dedicated railway communication network (F5G-R or railway data network). It adopts a data bus monitoring mode, listening to the railway freight yard operation data bus at the data link layer and subscribing to the "yard / + / realtime" topic via the MQTT protocol to passively receive real-time operation data broadcast from adjacent tracks.

[0060] During operation, the monitoring unit in another section broadcasts real-time operation data at a frequency of 1Hz. The data format is JSON and includes the following fields: yard_id (yard / track number), container_type (box type identifier), angular_velocity_sequence (lifting posture angular velocity sequence, including X / Y / Z three-axis angular velocities, sampling frequency 50Hz, duration 2 seconds before and after lifting off the seat / ground), and ftr_lock_abnormal (FTR lock engagement abnormality flag, boolean value).

[0061] The similarity indexing process employs the Locality Sensitive Hash (LSH) algorithm: mapping the high-dimensional attitude angular velocity sequence (3 axes × 200 sampling points = 600 dimensions) into a 128-bit hash signature. Specifically, the attitude angular velocity sequence is subjected to Fast Fourier Transform (FFT) to extract frequency domain features (the first 20 frequency points). The frequency domain feature vector (60 dimensions) is then input into a trained family of LSH functions (constructed based on a p-stable distribution, with a hash bucket width w = 0.5 and a hash table size L = 10) to generate a 128-bit signature. An inverted index is built to achieve fast similarity bin retrieval, using the bin identifier as the primary key and the hash signature as the index key to store the corresponding attitude angular velocity distribution (including the mean curve and variance band).

[0062] When the container type identifiers of the target double containers in this loading and unloading line operation match, the corresponding attitude angular velocity distribution is extracted as the pre-rehearsal reference benchmark for this container operation. During matching, the exact match between the target container type identifier and the index primary key is calculated; if the match is successful, all hash signatures under that primary key are extracted, the Hamming distance with the real-time sequence signature of the current operation is calculated, and the three records with the smallest distance are selected, and the mean curve of their attitude angular velocity distribution is extracted as the pre-rehearsal reference benchmark.

[0063] The risk prediction unit is deployed on the computing node of the railway bureau's data center and interfaces with the railway freight station management system, the railway bureau's historical operation data center, and meteorological stations along the railway line through standard API interfaces.

[0064] The specific implementation of risk score calculation:

[0065] For each container slot in the train formation information (identified by a two-dimensional coordinate system of car number and container slot), calculate the risk score R:

[0066] R = α·f history +β·S wind +γ·ρ cargo ;

[0067] In the formula: f history To determine the frequency of off-center loading at this container location in historical train operations, analyze all operational records for this train or loading / unloading line combination over the past 24 months, and calculate the ratio of off-center loading alarms to total operational operations; S wind This represents the wind pressure exposure area at the train location where the container is situated. The location of the container in an open area, cutting, or bridge section is calculated based on the train's operating line and freight yard location, and is calculated according to the actual wind-exposed area. The station's sheltered area is set to 0. ρ cargo The abnormality of the weight-to-volume ratio of the cargo in this container is calculated using the formula ρ. cargo =|W cargo / V cargo -ρ avg | / σ ρ W cargoFor cargo weight, V cargo For the volume of goods, ρ avg σ is the average weight-to-volume ratio of similar goods. ρ α represents the standard deviation; α, β, and γ are weighting coefficients, which are set to 0.5, 0.3, and 0.2 in this embodiment.

[0068] Generation of the off-center load risk heatmap: Using the train formation plan (or freight yard loading and unloading line plan) as the base map, the risk score R of each container position is mapped to a color intensity (R < 0.3 is green for low risk, 0.3 ≤ R < 0.7 is yellow for medium risk, and R ≥ 0.7 is red for high risk), generating a two-dimensional heatmap. The high-risk container positions (container number list) with a risk score R ≥ 0.7 are pushed to the pre-loaded data cache pool via a message queue, triggering the enhanced monitoring flag of the corresponding pre-loaded work file (setting the flag is_enhanced=true).

[0069] (ii) The critical pre-screening module is a critical pre-screening unit, which is an embedded controller deployed at the end of the spreader of the railway container loading and unloading equipment (gantry crane or reach stacker). It is connected to the spreader locking pin position sensor, micro-motion actuator and inertial response acquisition device via CAN bus.

[0070] Acquisition of the locking pin insertion depth signal: Four inductive displacement sensors (four sets per box, eight sets in total for both boxes) are arranged on the inner wall of the FTR lock hole to detect the locking pin insertion depth. When all eight sensors detect a metal proximity signal and the depth value is within the range of 20-35mm (the standard locking depth is 28mm, with an allowable deviation of ±7mm), it is determined that the locking pins of both boxes have entered the lock hole but are not fully locked (the full locking depth is 35mm, achieved by a mechanical locking groove).

[0071] Trial micro-disturbance application: Trial micro-disturbances are applied independently to the two containers. The micro-actuators include a vertical actuator (servo electric cylinder, stroke ±10mm, accuracy 0.1mm), a horizontal actuator (servo electric cylinder, stroke ±6mm, accuracy 0.05mm), and a rotary actuator (servo motor + reducer, rotation angle ±1°, accuracy 0.01°). The control algorithm is as follows:

[0072] Vertical displacement: z(t) = 5·sin(2π·0.5·t);

[0073] Horizontal displacement: x(t) = 3·sin(2π·0.3·t);

[0074] Rotation angle: θ(t) = 0.5·sin(2π·0.2·t);

[0075] Where t is time, in seconds, and the duration is 10 seconds (5 vertical cycles, 3 horizontal cycles, and 2 rotational cycles). The motion of the three degrees of freedom is superimposed to form a spatial composite micro-perturbation.

[0076] Timing control of independent micro-perturbations in two boxes: To avoid mutual interference between the micro-perturbations of the two boxes, a staggered application strategy is adopted: 1) Micro-perturbation of box A: t∈[0, 10] seconds; 2) Micro-perturbation of box B: t∈[12, 22] seconds; 3) A 2-second interval is used for vibration attenuation.

[0077] Acceleration response sequence acquisition: The inertial response acquisition unit consists of two sets of triaxial accelerometers deployed at the four corners of the spreader (one set per box, totaling two sets, range ±2g, resolution 0.001g, sampling frequency 100Hz). During the application of micro-perturbations, triaxial acceleration data from four measuring points in each box are acquired simultaneously to form a dual-box acceleration response sequence with dimensions of 2 boxes × 4 measuring points × 3 axes × 1000 sampling points (10 seconds × 100Hz).

[0078] Track baseline modal subtraction processing: Before modal parameter identification, the track-vehicle coupled vibration baseline signal is collected using triaxial accelerometers deployed on the railway flatcar bogies or car body underframe (or using historical calibration data). FFT analysis is performed on the baseline signal to extract characteristic frequency components (typically 20-80Hz) caused by rail short-wave irregularities (wavelength 0.5-3m) and track bed elastic deformation. Adaptive filtering (LMS algorithm or frequency domain notch filtering) is applied to the container acceleration response sequence to eliminate baseline interference components.

[0079] Inter-box coupling mode identification processing: 1) Data preprocessing: Detrending and bandpass filtering (passband 0.1-10Hz, removing high-frequency noise and DC drift) are performed on the dual-box acceleration response sequences after baseline subtraction; 2) Single-box mode identification: Hankel matrix is ​​constructed for each box data, and the covariance-driven random subspace algorithm (SSI) is used to identify the single-box mode parameters and extract the single-box natural frequency f. n A f n B And damping ratio ζ n A ζ n B 3) Inter-cell coupling mode identification: Construct a Hankel matrix by combining the data from both cells, identify the joint modal parameters of the two cells, and extract the joint natural frequency f of the two cells. n couple and combined damping ratio ζ n couple 4) Calculation of phase difference between the two cells: The phase difference φ between the two cells at the main modal frequencies is calculated by cross-power spectrum analysis. AB5) Feature extraction and determination: Calculate the dual-box coupling frequency offset: Δf couple =Σ|(f n couple -f n0 couple ) / f n0 couple | / 3; Determine the phase difference between the boxes: |φ A -φ B |;When Δf couple >0.20 or |φ A -φ B If the angle is greater than 45°, it is determined that the loading between containers is not coordinated, and the ABORT command is output to terminate the operation; otherwise, a qualified pre-judgment result is output, and the process can proceed to the next step.

[0080] (III) The multi-level interception module is a multi-level interception and detection unit deployed along the descent path of the spreader of the railway container loading and unloading equipment, such as... Figure 2 As shown.

[0081] The spatial locations of the three detection points are as follows: 1) First-level mechanical alignment detection point: located 500mm above the descent path of the spreader (when the locking pin is about to contact the locking hole); 2) Second-level locking pin engagement detection point: located 200mm above the descent path of the spreader (during the process of the locking pin being inserted into the locking hole); 3) Third-level off-seat / off-ground posture detection point: located 50mm below the descent path of the spreader (when the container is about to be off-seat or off the ground).

[0082] The first-level mechanical alignment detection points measure the height difference between the lifting device and the top corner of the two boxes using laser rangefinder sensors (0.1mm accuracy, 8 measuring points in total for both boxes) deployed at the four corners of the lifting device. The mechanical alignment deviation value (horizontal offset in the X / Y direction) of the FTR lock of each box is calculated. When the horizontal offset of either box exceeds ±5mm, the first-level abnormal signal is output. The second-level locking pin engagement detection points measure the engagement depth value (obtained by the locking pin displacement sensor) and engagement uniformity index using strain gauge pressure sensors (range 0-50kN, accuracy 0.5%FS, 8 sets of sensors in total for both boxes) deployed at the root of the locking pin.

[0083] Meshing uniformity index calculation: For each box with 4 sets of locking pins under force, calculate the ratio of standard deviation to mean: U = std(F) i ) / mean(F i When the engagement depth of any gearbox is less than 20mm (safe depth) or the engagement uniformity index is greater than 0.15 (non-uniformity threshold), a second-level abnormal signal is output.

[0084] The third-level off-seat / off-ground attitude detection point uses a dual-axis tilt sensor (accuracy 0.01°, one set for each of the two boxes) deployed at the center of the spreader to measure the angular acceleration value of each box at the instant of off-seat or off-ground (obtained by second differentiation of the tilt signal, sampling frequency 100Hz, taking the maximum value within a 0.5-second window before and after off-seat / off-ground). When the angular acceleration value of any box exceeds 0.5° / s² (dynamic imbalance threshold), a third-level abnormal signal is output.

[0085] Progressive judgment logic: During system initialization, only the first-level mechanical alignment detection point is activated. If the first level determines that both boxes are normal, the second-level locking pin engagement detection point is activated. If the second level determines that both boxes are normal, the third-level off-seat / off-ground posture detection point is activated. If any level determines that either box is abnormal, a path locking command is generated. This command cuts off the lifting device descent control circuit via a hard-wired circuit, prohibiting further operation. After manual confirmation that the abnormality has been eliminated, all subsequent checks at that level must be re-executed starting from the abnormality level.

[0086] (iv) The active control module is an active correction control unit, which is deployed in the IPC of the railway freight yard control building or station dispatching room and is connected to the hydraulic leveling mechanism of the loading and unloading equipment lifting device via EtherCAT bus.

[0087] Control strategy switching logic: Receive dual-box modal parameters (single-box natural frequency offset Δf) from the critical pre-screening unit. A , Δf B Dual-box coupling frequency offset Δf couple Inter-box phase difference φ AB ) or dual-box takeoff / takeoff attitude data (angular acceleration value α) from a multi-level interception and detection unit. A α B ), and the inter-box centroid coupling risk integral R of the pre-monitoring module. couple When any single box Δf > 0.10 or |α| > 0.3° / s² or R couple When the value is ≥1.0, switch to active correction mode.

[0088] Single-box center of gravity offset calculation: Based on the load values ​​of the four corner load cells of each box (A front left, B front right, C rear left, D rear right), calculate the center of gravity offset coordinates of both boxes respectively:

[0089] Box 1 (i=1):

[0090] X offset 1 =(B1+D1-A1-C1)·Lx / (A1+B1+C1+D1);

[0091] Y offset 1=(A1+B1-C1-D1)·Ly / (A1+B1+C1+D1);

[0092] Box 2 (i=2):

[0093] X offset 2 =(B2+D2-A2-C2)·Lx / (A2+B2+C2+D2);

[0094] Y offset 2 =(A2+B2-C2-D2)·Ly / (A2+B2+C2+D2);

[0095] In the formula, Lx is the transverse half-length of the container (1512mm for a 20-foot container and 3024mm for a 40-foot container), and Ly is the longitudinal half-width (1219mm).

[0096] Overall center of gravity shift of the flatcar: The center of gravity shift of the two individual boxes is weighted and calculated to obtain the overall center of gravity shift of the flatcar.

[0097] X total =(W1·X offset 1 +W2·X offset 2 ) / (W1+W2);

[0098] Y total =(W1·Y offset 1 +W2·Y offset 2 ) / (W1+W2);

[0099] W1 and W2 represent the weight of the two boxes.

[0100] Dual-box collaborative correction determination: 1) When |X total |>ΔX max (Flatcar center of gravity threshold) or |W1-W2|>ΔW max When the weight difference threshold is reached, the dual-box collaborative correction mode is triggered; 2) General threshold: ΔX max =150mm, ΔW max = 5 tons (BX70B type: ΔX) max =100mm, ΔW max =3 tons).

[0101] Calculation of target correction torque:

[0102] Single-box target correction torque:

[0103] Mx i =Kp·X offseti +Kd·d(X offset i ) / dt;

[0104] My i =Kp·Y offset i +Kd·d(Y offset i ) / dt;

[0105] In the formula, the proportionality coefficient Kp = 500 N·m / mm, the differential coefficient Kd = 50 N·m·s / mm, and the risk integral R between the center of gravity of the boxes is... couple Positive correlation, when R couple When the value is ≥1.0, Kp and Kd increase by 20%.

[0106] Overall wheel load distribution constraints for flatcars: Coordinated correction of double-box alignment must meet the torque difference constraint of a single box.

[0107] ΔM=|M 1 -M 2 |<0.5·M total ;

[0108] In the formula, M total =Kp·X total +Kd·d(X total ) / dt represents the overall target correction torque of the flatcar.

[0109] If the torque difference of a single box does not meet the constraints, adjust the torque of each single box proportionally, prioritizing the correction of the overall center of gravity offset, while minimizing the imbalance of a single box.

[0110] Railway flatcar suspension system stiffness compensation: When a container is placed on the railway flatcar, the real-time suspension stiffness ks is obtained through pressure sensors deployed on the air / steel springs of the flatcar bogie. The additional attitude angle θs caused by the elastic deformation of the car body underframe is calculated, and the target correction torque is adjusted accordingly.

[0111] M'x i =Mx i -ks·θs·Lx;

[0112] M'y i =My i -ks·θs·Ly;

[0113] In the formula, θs is measured in real time by the tilt sensor of the vehicle body chassis.

[0114] The differential output of the four independent hydraulic cylinders of the lifting device's hydraulic leveling mechanism: The outputs FA, FB, FC, and FD of the four independent hydraulic cylinders (corresponding to the four corners of the lifting device) are based on the corrected target correction torque M'x. i Myi calculate:

[0115] FA=F0-M'y / (2·Ly)+M'x / (2·Lx);

[0116] FB=F0-M'y / (2·Ly)-M'x / (2·Lx);

[0117] FC=F0+M'y / (2·Ly)+M'x / (2·Lx);

[0118] FD=F0+M'y / (2·Ly)-M'x / (2·Lx);

[0119] Wherein, F0 is the static load distribution caused by the weight of a single container, F0 = (A + B + C + D) / 4. The hydraulic cylinder controls the output through a proportional valve, with a response time of <100ms, and continues to apply force until the container is lifted off the ground or the center of gravity deviation drops below the threshold.

[0120] (v) Optional Modules

[0121] Internal and external sensor fusion unit: The passive RFID pressure sensing tag adopts the UHF band 860-960MHz and transmits pressure field data with backscatter modulation; the stress and strain sensor adopts PZT-5H piezoelectric ceramic self-powered to collect three-dimensional force state data; the data spatiotemporal alignment processing uses the container number as the associated primary key, the system unified clock as the reference, and ±100ms as the time window matching rule to establish a container-lock coupling force model.

[0122] Passive RFID pressure sensing tags: These UHF band (860-960MHz) passive RFID tags integrate pressure-sensitive elements (capacitive pressure sensors, range 0-100kPa, resolution 0.1kPa), deployed at the four corners and center of the bottom surface inside each container (5 measuring points per container, 10 measuring points for two containers). The tags transmit pressure data back to the RFID reader at the end of the spreader via backscatter modulation (transmission power 30dBm, reading distance 0.5-3m), requiring no built-in battery. During the lifting process of the spreader on the railway container handling equipment, when the spreader approaches the top of the container (distance <1m), the RFID reader activates the tag, collecting pressure field data at a frequency of 10Hz.

[0123] Stress-strain sensors: MEMS piezoelectric strain sensors are embedded in the stress concentration areas inside each FTR locking pin (8 sets of locking pins in two boxes, with one sensor embedded in each set of pins). Self-powered operation and signal output are achieved through the piezoelectric effect of the locking pin material (PZT-5H piezoelectric ceramic, piezoelectric coefficient d33≈593pC / N). The sensors collect three-dimensional (axial, shear, and bending) stress data of the locking pins at a sampling frequency of 100Hz.

[0124] Data spatiotemporal alignment processing specifically includes:

[0125] Timestamp alignment: The time reference is T0, which is the trigger time of the lifting action of the lifting device (determined by the lifting device height encoder, triggered when the lifting device rises from rest at a speed >0.1m / s). The timestamp for the pressure field data is the RFID reader reception time T1, and the timestamp for the three-dimensional stress state data is the stress-strain sensor sampling time T2. The alignment rule is: when |T1-T0|<100ms and |T2-T0|<100ms, the data are considered to be from the same acquisition period.

[0126] Spatial Alignment: A three-dimensional coordinate system is established with the geometric center of each container as the spatial reference origin O (0, 0, 0) (X-axis along the length of the container, Y-axis along the width, and Z-axis along the height). The coordinates of the measurement points for the pressure field data are preset fixed values ​​(the four corners and the center); the coordinates of the locking pins for the three-dimensional stress state data are determined according to the FTR lock installation position (the coordinates of the four corner locking pins). Through coordinate transformation, the two types of data are mapped to a unified spatial reference frame.

[0127] Coupled force model generator: Receives aligned pressure field data (pressure values ​​Pi at 5 measuring points per container) and triaxial force state data (triaxial forces Fjx, Fjy, Fjz at 4 locking pins per container), and establishes a container-locking coupled force model for each container.

[0128] Internal load distribution: The pressure distribution field P(x, y) on the bottom surface of the container is reconstructed using the pressure values ​​Pi at 5 measuring points and a bilinear interpolation algorithm.

[0129] External lock force: The three-dimensional forces of the four locking pins are combined to form the total lock force, Flock.

[0130] Coupled constraint equations: Based on the static equilibrium conditions, establish the following system of equations:

[0131] ΣFlock=∫∫P(x,y)dxdy+W container ;

[0132] ΣMlock=∫∫P(x,y)·r(x,y)dxdy+M container ;

[0133] In the formula, W container For the container's own weight, M container Let r(x, y) be the torque caused by its own weight, and r(x, y) be the position vector.

[0134] The equation residuals are solved by the least squares method. When the residuals exceed the threshold (indicating a mismatch between the internal off-center load and the external force), a mismatch mode alarm is output.

[0135] Environmental adaptive threshold unit: IPC deployed in the railway freight yard control building or station dispatching room, which connects to the railway engineering section track monitoring system, railway meteorological stations and train marshalling information system through standard interfaces.

[0136] Construction of the dynamic threshold correction matrix: The matrix dimension is 3×4×3, where:

[0137] First dimension (track stiffness level): Ballastless track / bridge section (stiffness > 200kN / mm), ballasted track straight section (100-200kN / mm), ballasted track turnout area / soft soil subgrade (< 100kN / mm).

[0138] The second dimension (wind zone level range): sheltered stations (wind speed < 6 m / s), general open areas (6-12 m / s), strong wind areas / bridges / road cuts (12-18 m / s), extreme wind areas (> 18 m / s);

[0139] The third dimension (total weight range of train formation): light-load trains (total weight < 1000t), medium-load trains (1000-3000t), and heavy-load trains (> 3000t).

[0140] The matrix elements store the bias warning threshold correction coefficient kp and the bias warning threshold correction coefficient kw.

[0141] Threshold correction calculation: Multiply the basic threshold (horizontal ±100mm, vertical ±50mm, single box weight 30480kg, weight difference between two boxes 5 tons) by the correction factor to obtain the corrected off-center load warning threshold and off-center weight warning threshold.

[0142] Example 2:

[0143] This embodiment focuses on the "one wagon, two containers" China-Europe freight train transport operation using the BX70B type container flatcar (26.366m in length, 68t in load capacity, capable of carrying two 40-foot containers simultaneously) at the Shaliang Logistics Park of China Railway Hohhot Bureau Group Co., Ltd.

[0144] Taking the first BX70B China-Europe freight train on April 8, 2025 as an example, flatcar 1 was loaded with container A (refrigerator, cargo weight 18t) and container B (television, cargo weight 15t).

[0145] (a) Pre-monitoring module

[0146] The preloaded data cache pool receives BX70B flatcar double-box grouping information from TMIS, including: 1) Flatcar number: BX70B-001; 2) Double-box numbers: Box A (TBJU1234567), Box B (TBJU7654321); 3) Double-box cargo type codes: Box A (refrigerated), Box B (general dry goods); 4) Double-box weight difference warning threshold: 3 tons; 5) Composite center of gravity offset threshold: 100mm.

[0147] Inter-box centroid coupling risk integral R couple calculate:

[0148] R couple =|W1-W2| / 3+|X1+X2| / 0.2;

[0149] In this context, W1 and W2 are units of tons, and X1 and X2 are units of meters.

[0150] Taking container A (18t) and container B (15t) as examples: 1) Assume the estimated values ​​of the center of gravity offset of a single container: X1 = 0.02m (2cm to the right for container A), X2 = 0.03m (3cm to the right for container B); 2) R couple =|18-15| / 3+|0.02+0.03| / 0.2=1.0+0.25=1.25; 3) R couple =1.25≥1.0, triggering the dual-box collaborative enhancement monitoring flag (is_dual_enhanced=true).

[0151] (II) Critical Pre-screening Module

[0152] Trial micro-perturbations were applied independently to two 40-foot containers loaded on a BX70B flatcar, and acceleration response sequences of the two containers were collected.

[0153] Dual-enclosure coupling mode identification: 1) Dual-enclosure coupling frequency offset threshold: 0.20; 2) Allowable range of inter-enclosure phase difference: 45 degrees; 3) When the dual-enclosure coupling frequency offset Δf couple >0.20 or inter-cell phase difference |φ A -φ B When the angle is greater than 45°, it is determined that the loading between containers is not coordinated, and the ABORT command is output to terminate the operation.

[0154] (III) Multi-level interception module

[0155] In the "one car, two boxes" working condition of the BX70B flatcar, each box has 4 sets of FTR locks, for a total of 8 sets of FTR locks for both boxes.

[0156] The second-stage locking pin engagement detection point receives eight sets of FTR locking engagement depth values ​​and engagement uniformity indices from the locking pin pressure sensor.

[0157] (iv) Active control module

[0158] Calculation of the combined center of gravity of the two boxes: 1) Overall center of gravity offset threshold of the flatcar: 100mm; 2) Weight difference threshold between the two boxes: 3 tons;

[0159] Taking container A (18t, X1=0.02m) and container B (15t, X2=0.03m) as examples:

[0160] X total =(18×0.02+15×0.03) / (18+15)=(0.36+0.45) / 33=0.0245m=24.5mm;

[0161] |X total |=24.5mm<100mm, the composite center of gravity is within the limit.

[0162] |W1-W2|=|18-15|=3t=3t, the weight difference reaches the threshold boundary, triggering a yellow warning.

[0163] If X1 = 0.05m and X2 = 0.06m:

[0164] X total =(18×0.05+15×0.06) / 33=(0.9+0.9) / 33=0.0545m=54.5mm<100mm;

[0165] |W1-W2|=3t=3t;

[0166] Although the center of gravity of the synthesis did not exceed the limit at this time, the single box was heavily off-center, and the system still triggered the dual box collaborative enhanced monitoring.

[0167] Calculation of target correction torque:

[0168] Kp=500 N·m / mm, Kd=50 N·m·s / mm;

[0169] Since R_couple = 1.25 ≥ 1.0, Kp and Kd increase by 20%.

[0170] Kp'=600 N·m / mm, Kd'=60 N·m·s / mm;

[0171] M total =Kp'·X total +Kd'·d(X total ) / dt;

[0172] Single-box moment difference constraint: ΔM=|M1-M2|<0.5·M total .

[0173] Example 3:

[0174] The container FTR lock operation and off-center loading / weight monitoring and early warning method for the "one car, two containers" transportation scenario of railway flatcars in this embodiment are based on the system execution of Embodiment 1, such as... Figure 3 As shown, it includes the following steps:

[0175] S1. Pre-monitoring steps: The pre-loaded data cache pool receives multi-source data at time T before the container arrives, performs dual-container joint pre-association processing to generate dual-container joint pre-loading operation files; the cross-domain attitude database listens to broadcast data from adjacent tracks, performs LSH similarity indexing to establish similar container type attitude data and sends it to the pre-loaded data cache pool; the risk prediction unit performs risk integral calculation to generate an off-center load risk heatmap, and pushes high-risk indicators to the pre-loaded data cache pool; the inter-container center of gravity coupling risk integral R is executed. couple Calculate when R couple When the value is ≥1.0, the dual-box collaborative enhanced monitoring marker is triggered.

[0176] Calculation of time T: Receive the estimated arrival and departure times T of the train from the Railway Station Freight Management System (TMIS). TMIS (In UTC timestamp format), query the historical statistics database to obtain the average arrival time Δt for this train type. arrival (Typically 10-20 minutes), calculate the estimated arrival time T at the loading / unloading line. arrival =T TMIS +Δt arrival The processing time Δt is estimated based on the amount of data in the preloaded job file. load (Approximately 2 minutes per thousand records), set a safety margin Δt margin =10 minutes, then T=T TMIS -Δt load -Δt margin .

[0177] S2. Critical Pre-screening Step: The critical pre-screening unit receives the insertion depth signal of the dual-box locking pin (8-channel analog input of inductive sensors, range 4-20mA corresponding to a depth of 0-50mm). When the depth value of all channels is ∈ [20, 35]mm, it is determined that the dual boxes have entered the critical window period, triggering the micro-motion actuator.

[0178] Three-degree-of-freedom micro-perturbations were applied independently to the two boxes (staggered timing: box A [0, 10] s, box B [12, 22] s), with parameters as shown in Example 1. The motion lasted for 10 seconds, during which the inertial response acquisition unit collected acceleration data at a frequency of 100 Hz.

[0179] Perform orbital baseline mode subtraction and SSI mode parameter identification, and determine inter-cell coupling (Δf). couple ≤0.20 and |φ A -φ B|≤45°).

[0180] When the condition is deemed unqualified, an "ABORT" command is sent to the loading and unloading equipment control system to terminate the current operation and prompt the operator to check the loading status of the two containers; when the condition is deemed qualified, a "PROCEED" command is sent to allow the operation to proceed to the next step.

[0181] S3. Multi-level interception steps: The multi-level interception detection unit activates the three-level detection points sequentially according to the descent path of the spreader, simultaneously detecting both boxes:

[0182] The first level of mechanical alignment detection points: The height of the four corners of the two boxes is measured by a laser rangefinder (a total of 8 measuring points), and the horizontal offset of the two boxes is calculated. When the offset of any box is greater than 5mm, a path locking command is generated.

[0183] Second-level locking pin engagement detection point: The force Fi (i=1, ..., 8) is measured by 8 sets of locking pin pressure sensors, and the engagement depth Di and engagement uniformity index U for each box are calculated. When any Di < 20mm or U > 0.15 for any box, a path locking command is generated.

[0184] The third-level off-seat / off-ground attitude detection point: Angular acceleration α = d²θ / dt² is measured by dual-box tilt sensors. When |α| > 0.5° / s² for either box, a path lock command is generated, triggering emergency braking.

[0185] Progressive judgment: After both boxes pass the judgment at each level, the next level of detection is activated; if any box fails the judgment, the failed box number and level are recorded, and after manual confirmation, the process restarts from that level.

[0186] S4. Active control step: The active correction control unit receives the dual-box mode parameters (Δf) from step S2. A , Δf B , Δf couple φ AB ) or dual-box off-seat / off-ground attitude data from step S3 (α) A α B When any single box Δf > 0.10 or |α| > 0.3° / s² or R couple When the value is ≥1.0, switch to dual-box collaborative correction mode.

[0187] Calculate the center-of-gravity offset (X) of each of the two boxes. offset 1 Y offset 1 ), (X offset 2 Y offset 2 The overall center of gravity of the composite flatcar shifts by X. total .

[0188] When |X total When |>100mm (BX70B type) or |W1-W2|>3t (BX70B type), dual-box coordinated correction is triggered.

[0189] Calculate the target correction torque M total =Kp·X total +Kd·d(X total ) / dt, query the inter-box centroid coupling risk integral R couple If R couple If ≥1.0, Kp and Kd increase by 20%.

[0190] The moment difference of the single-box constraint is ΔM = |M1 - M2| < 0.5·M total .

[0191] When seated on the railway flatcar, the bogie suspension stiffness ks and the car body underframe attitude angle θs are queried, and the target correction torque is corrected to M'x and M'y according to the formula in Example 1.

[0192] Control current (4-20mA corresponding to 0-100% output) is sent to the proportional valves of the four hydraulic cylinders via EtherCAT bus. The feedback from the cylinder pressure sensor is monitored in real time, and closed-loop adjustment is performed to achieve the target output. The center of gravity offset of the two containers is continuously monitored. When the offset of both containers drops below the threshold or the containers are completely off-seat / off the ground (height > 100mm), the dual-container collaborative correction mode is exited.

[0193] The critical pre-screening step S2 and the multi-level interception step S3 are complementary in timing: the critical pre-screening step S2 completes the risk pre-screening of the two boxes during the first critical window period (lock pin insertion is not locked), and the multi-level interception step S3 completes the multi-level interception of the two boxes during the second critical window period (spreader descent and leaving the seat / ground). The termination decision of S2 takes precedence over the execution of S3.

[0194] S5, Sensor Fusion Step (Optional Enhancement): When the system is equipped with internal and external sensor fusion units, the following steps are performed on both boxes during the lifting process:

[0195] Passive RFID pressure sensing tag activation: The RFID reader at the end of the lifting device transmits a carrier signal (frequency 920-925MHz), and the dual-box tags receive the energy and return pressure data. Reading distance: 0.5-3m, reading success rate: >99%.

[0196] The stress-strain sensor is self-powered: the deformation of the eight sets of locking pins in the dual-box system generates piezoelectric charges, which are converted into voltage signals by the charge amplifier to collect the three-dimensional force.

[0197] Data spatiotemporal alignment: Perform timestamp alignment (±100ms window) and spatial alignment (coordinate transformation).

[0198] Coupled force model establishment: Solve the coupling constraint equations of the two containers separately and calculate the residuals. When the residual of any container is greater than the threshold (set to 5% of the container weight in this embodiment), a mismatch alarm is output, indicating that the cargo inside the container has shifted or the lashing has failed.

[0199] S6. Environmental Adaptation Steps (Optional Enhancement): When the system is configured with an environmental adaptive threshold unit, this step is executed upon system startup or when weather / orbital conditions change.

[0200] Environmental parameter collection: Query the railway maintenance section's track monitoring system to obtain the current track stiffness coefficient of the loading and unloading line, query the meteorological stations along the railway line to obtain the current wind speed and direction, and query the train formation information system to obtain the current total weight of the train formation.

[0201] Threshold Correction: Query the dynamic threshold correction matrix to obtain the correction coefficients kp and kw, calculate the corrected dual-box warning threshold, and update it to the system configuration.

[0202] In the above embodiments, the descriptions of each embodiment have different focuses. For parts that are not described in detail or recorded in a certain embodiment, please refer to the relevant descriptions of other embodiments.

[0203] In various embodiments, the hardware implementation of the technology can directly utilize existing smart devices, including but not limited to industrial control computers, PCs, smartphones, handheld devices, and floor-standing devices. Its input device preferably uses an on-screen keyboard, its data storage and computing modules utilize existing memory, calculators, and controllers, its internal communication modules utilize existing communication ports and protocols, and its remote communication utilizes existing GPRS networks, the World Wide Web, etc.

[0204] Those skilled in the art will clearly understand that, for the sake of convenience and brevity, the above-described division of functional units and modules is merely an example. In practical applications, the above functions can be assigned to different functional units and modules as needed, that is, the internal structure of the device can be divided into different functional units or modules to complete all or part of the functions described above. The functional units and modules in the embodiments can be integrated into one processing unit, or each unit can exist physically separately, or two or more units can be integrated into one unit. The integrated unit can be implemented in hardware or as a software functional unit. Furthermore, the specific names of the functional units and modules are only for easy differentiation and are not intended to limit the scope of protection of this application. The specific working process of the units and modules in the above system can be referred to the corresponding process in the foregoing method embodiments, and will not be repeated here.

[0205] In the embodiments provided by this invention, it should be understood that the disclosed apparatus / terminal devices and methods can be implemented in other ways. For example, the apparatus / terminal device embodiments described above are merely illustrative. For instance, the division of modules or units is only a logical functional division, and in actual implementation, there may be other division methods. For example, multiple units or components may be combined or integrated into another system, or some features may be ignored or not executed. Furthermore, the shown or discussed mutual couplings or direct couplings or communication connections may be through some interfaces; indirect couplings or communication connections between devices or units may be electrical, mechanical, or other forms. Units described as separate components may or may not be physically separate, and components shown as units may or may not be physical units, i.e., they may be located in one place or distributed across multiple network units. Some or all of the units can be selected to achieve the purpose of this embodiment according to actual needs.

[0206] In the various embodiments of the present invention, the functional units can be integrated into one processing unit, or each unit can exist physically separately, or two or more units can be integrated into one unit. The integrated units can be implemented in hardware or as software functional units. If the integrated module / unit is implemented as a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the present invention can implement all or part of the processes in the methods of the above embodiments, or it can be accomplished by a computer program instructing related hardware. The computer program can be stored in a computer-readable storage medium, and when executed by a processor, it can implement the steps of the various method embodiments described above. The computer program includes computer program code, which can be in the form of source code, object code, executable files, or certain intermediate forms. The computer-readable medium can include: any entity or device capable of carrying computer program code, recording media, USB flash drives, portable hard drives, magnetic disks, optical disks, computer memory, read-only memory (ROM), random access memory (RAM), electrical carrier signals, telecommunication signals, and software distribution media, etc.

Claims

1. A railway container FTR lock operation and off-center loading / weight monitoring and early warning system, characterized in that, It includes a pre-monitoring module, a critical pre-screening module, a multi-level interception module, and an active control module; The pre-monitoring module is used to generate pre-monitoring information based on historical operation data, cross-domain operation data and environmental risk data before the container arrives at the railway freight yard operation area, and push the pre-monitoring information to the downstream module; The pre-monitoring module receives double-container grouping information from the train grouping information system, generates a double-container joint pre-loading operation file, and performs inter-container center of gravity coupling risk integral calculation on two containers loaded on the same flatcar. When the inter-container center of gravity coupling risk integral exceeds the threshold, a double-container collaborative enhanced monitoring marker is triggered. The critical pre-screening module is connected to the pre-monitoring module and is used to perform stability prediction on two containers based on the pre-monitoring information when the locking pin of the spreader of the railway container loading and unloading equipment is inserted into the FTR lock hole but not completely locked, and output the prediction result. The multi-level interception module is connected to the critical pre-screening module and is used to progressively detect the FTR lock operation status and container attitude of the two containers along the descent path of the spreader of the railway container loading and unloading equipment when the prediction result is stable. If any detection point is abnormal, the operation path is locked. The active control module is connected to the critical pre-screening module and the multi-level interception module respectively. It is used to receive the pre-judgment results or the off-seat / off-ground posture data of the progressive detection, calculate the single-container center of gravity offset of the two containers respectively and synthesize the overall center of gravity offset of the flatcar. When the overall center of gravity offset of the flatcar exceeds the flatcar center of gravity threshold or the weight difference between the two containers exceeds the weight difference threshold, the dual-container collaborative correction mode is triggered, and the target correction torque is applied so that the two containers reach the critical balance state at the moment of lifting or sitting.

2. The monitoring and early warning system according to claim 1, characterized in that, The pre-monitoring module includes a pre-loaded data cache pool, a cross-domain attitude database, and a risk prediction unit. The pre-loaded data cache pool receives double-box formation information from the railway freight station management system, historical center of gravity offset records from the historical database, and lock wear parameters from the FTR lock status monitoring system to generate a double-box joint pre-loading operation file. The cross-domain attitude database receives real-time operation data broadcast from monitoring units in other sections of adjacent tracks through a data bus listening mode to establish similar box type attitude data with box type identifier as the primary key. The risk prediction unit receives train formation information, historical train eccentricity data, and wind zone weather forecast information to output an eccentricity risk heat map.

3. The monitoring and warning system of claim 1, wherein, The critical pre-screening module independently applies tentative micro-disturbances to the two containers, collects the acceleration response sequence of the two containers, and extracts the joint natural frequency and damping ratio of the two containers through inter-container coupling mode identification. When the offset of the coupling frequency of the two containers exceeds the threshold or the phase difference between the containers exceeds the allowable range, it is determined that the loading between the containers is not coordinated, outputs the stability failure prediction result, and terminates the current operation process. The modal parameter extraction of the tentative micro-disturbance also includes a track baseline mode subtraction step: before identifying the modal parameters of the container-spreader system, the track-vehicle coupled vibration baseline signal under the empty state of the railway flatcar is first collected, and the interference of rail irregularities and track bed elastic deformation on the acceleration response sequence is eliminated by frequency domain subtraction or adaptive filtering.

4. The monitoring and warning system of claim 1, wherein, The calculation of the target correction torque also incorporates the stiffness compensation coefficient of the railway flatcar suspension system: when the container is placed on the railway flatcar, the differential output of the hydraulic cylinder is corrected based on the real-time stiffness feedback of the air spring / steel spring of the vehicle bogie to compensate for the attitude error caused by the elastic deformation of the car body frame.

5. The monitoring and warning system of claim 1, wherein, The multi-level interception module is a multi-level interception detection unit deployed along the descent path of the spreader of the railway container loading and unloading equipment, including a first-level mechanical alignment detection point, a second-level locking pin engagement detection point, and a third-level locking pin engagement detection point. The first-level mechanical alignment detection point receives the mechanical alignment deviation value of the dual-box FTR lock from the spreader vision alignment system; the second-level lock pin engagement detection point receives the engagement depth value and engagement uniformity index of the dual-box FTR lock from the lock pin pressure sensor; the third-level off-seat / off-ground posture detection point receives the angular acceleration value of the dual-box at the moment of off-seat or off-ground from the spreader tilt angle sensor; the multi-level interception detection unit generates a path locking command when it receives any level of abnormal signal.

6. The monitoring and warning system of claim 1, wherein, It also includes an internal and external sensor fusion unit; the internal and external sensor fusion unit includes: a passive RFID pressure sensing tag deployed on the bottom surface inside the container to collect data on the pressure field of the cargo distribution inside the container; a stress and strain sensor embedded inside the FTR lock pin to collect data on the three-dimensional force state of the lock pin; and a coupled force model generator to establish a container-lock coupled force model and identify the mismatch pattern between the internal cargo off-center load and the external lock force.

7. The monitoring and warning system of claim 1, wherein It also includes an environment-adaptive threshold unit; the environment-adaptive threshold unit receives track stiffness coefficient from the track monitoring system of the railway engineering section, wind speed and wind direction data from meteorological stations along the railway line, and total weight and axle load distribution data of the train formation information system, queries the built-in dynamic threshold correction matrix, and outputs the corrected off-center load warning threshold and off-center weight warning threshold.

8. The monitoring and warning system of claim 1, wherein, The system is applicable to BX70B type container flatcars: the risk integral R between container centers of gravity coupling couple =|W1-W2| / 3+|X1+X2| / 0.2, where W1 and W2 are the weights of the two boxes, and X1 and X2 are the lateral offsets of the center of gravity of a single box; the threshold for the coupling frequency offset of the two boxes is 0.20, and the allowable range for the phase difference between the boxes is 45 degrees; the threshold for the center of gravity of the flatcar is 100 mm, and the threshold for the weight difference is 3 tons; the second-level locking pin engagement detection point of the multi-level interception module receives 8 sets of FTR lock engagement depth values ​​and engagement uniformity index.

9. A railway container FTR lock operation and unbalance load monitoring and early warning method, based on the monitoring and early warning system of any one of claims 1-8, characterized in that, Includes the following steps: The pre-monitoring steps involve generating pre-monitoring information based on historical operation data, cross-domain operation data, and environmental risk data before containers arrive at the railway freight yard operation area; receiving double container marshalling information and generating a double container joint pre-loading operation file; performing inter-container center of gravity coupling risk integral calculation, and triggering a double container collaborative enhanced monitoring marker when the threshold is exceeded. The critical pre-screening step involves making stability predictions for two containers in a critical state where the locking pin of the spreader of the railway container loading and unloading equipment is inserted into the FTR lock hole but not fully locked. The multi-level interception process involves progressively detecting the FTR lock operation status and container attitude of the two containers along the spreader descent path when the prediction result indicates stability. The active control step receives the predicted results or the off-seat / off-ground posture data detected progressively, calculates the single-container center of gravity offset of the two containers and synthesizes the overall center of gravity offset of the flatcar. When the overall center of gravity offset of the flatcar exceeds the flatcar center of gravity threshold or the weight difference between the two containers exceeds the weight difference threshold, the dual-container collaborative correction mode is triggered, and the target correction torque is applied so that the two containers reach the critical balance state at the moment of lifting or sitting.

10. The monitoring and alerting method of claim 9, wherein, The critical pre-screening step and the multi-level interception step complement each other in sequence: the critical pre-screening step completes risk pre-screening during the first critical window period when the spreader contacts the container but is not completely locked, and the multi-level interception step completes multi-level interception during the second critical window period when the spreader descends and lifts off the seat / lifts off the ground. The two steps cover different risk exposure periods in the operation process, and the termination decision of the critical pre-screening step takes precedence over the execution of the multi-level interception step.

Images