Intelligent barrier control system and method based on running state feedback

Abstract

The invention discloses an intelligent barrier control system and method based on running state feedback, belongs to the technical field of intelligent control, and aims to solve the problems that a traditional barrier is insufficient in authority authentication accuracy, rigid in running mode and lagged in fault handling. According to the method, a passing trigger signal is received, multi-mode identification data is collected, dynamic permission verification is completed, and a handrail driving and resetting instruction is generated to control handrail operation; synchronously monitoring railing operation data, constructing an operation data set, recognizing an abnormal state in combination with a normal operation feature library, and positioning fault types to generate a fault type list; updating the traffic level according to the traffic trigger signal frequency, and dynamically switching the operation mode in combination with the dominant fault type; and a control fault is identified, a temporary control instruction is generated, the residual life of a fault associated component is calculated, and early warning information is generated to complete fault early warning. Intelligent management and control of barrier operation adjustment and fault handling are achieved, and the passing efficiency, the equipment safety and the operation stability are remarkably improved.

Description

Technical Field

[0001] This invention relates to the field of intelligent control technology, and more specifically to an intelligent control system and method for barrier gates based on operational status feedback. Background Technology

[0002] In traffic control scenarios such as parking lots and residential community entrances, barrier gates are core equipment for ensuring traffic order. Their operational reliability and control effectiveness directly affect traffic efficiency and scene safety. Currently, the core operating parameters of existing barrier gates are mostly preset fixed values, making it difficult to flexibly adjust the response state according to real-time traffic flow. During peak hours, the response speed cannot match the traffic demand, often causing traffic congestion; during off-peak hours, the lack of targeted parameter adaptation leads to unnecessary energy consumption, resulting in poor overall operational economy. Furthermore, traditional barrier gate anomaly monitoring methods are relatively rudimentary, relying mostly on simple threshold judgments or regular manual inspections, making it difficult to quickly detect potential faults. Fault detection and handling are significantly delayed, easily disrupting normal traffic flow and potentially causing the fault to escalate, increasing maintenance costs and downtime. Simultaneously, existing technology lacks the ability to effectively predict the lifespan of barrier gate core components. Component maintenance is mostly reactive, relying on replacement after failure, without the ability to plan maintenance in advance. Sudden failure of critical components can easily cause equipment shutdown, further disrupting traffic order. Furthermore, some barrier gate mechanisms lack effective emergency control measures after a malfunction, making it difficult to prevent the malfunction from worsening while ensuring basic passage functionality. This exacerbates equipment wear and tear and increases traffic safety risks. Therefore, to overcome these limitations, this invention proposes an intelligent control system and method for barrier gates based on operational status feedback. Summary of the Invention

[0003] To address the shortcomings of existing technologies, the present invention aims to provide an intelligent control system and method for barrier gates based on operational status feedback. This system solves the problem of how to dynamically adapt the barrier gate's operating mode to traffic flow and fault status, promptly identify faults and generate targeted control commands, and predict the remaining lifespan of fault-related components for effective early warning, thereby ensuring traffic efficiency and equipment operational safety.

[0004] To achieve the above objectives, the present invention provides the following technical solution:

[0005] The intelligent control system for barrier gates based on operational status feedback includes a status perception module, an analysis and diagnosis module, a feedback control module, and an early warning and interaction module.

[0006] The status awareness module is used to collect multimodal recognition data when a passage trigger signal is received. By performing authentication feature extraction and permission rule base matching on the multimodal recognition data, dynamic permission verification is performed to determine whether to start the barrier gate operation. If so, based on the current operating mode of the barrier gate, the barrier gate control operation is executed, barrier drive commands and barrier reset commands are generated, the barrier operation is controlled, and the feedback monitoring mechanism is started.

[0007] The analysis and diagnosis module is used to collect railing operation data when the feedback monitoring mechanism is activated, build an operation dataset according to the operation cycle, identify abnormal operation status and locate abnormal operation characteristics by combining the normal operation feature library of the current operation mode; and classify and match the abnormal operation characteristics with the fault rule library to locate the target fault rule and generate a fault type list.

[0008] The feedback control module is used to update the current passage flow level of the barrier gate based on the frequency of the passage trigger signal, and to locate the dominant fault type from the fault type list, so as to dynamically switch the operating mode of the barrier gate in combination with the passage flow level.

[0009] The early warning interaction module is used to identify the controlled fault types in the fault type list, generate temporary control commands for each controlled fault type in sequence, and calculate the remaining life prediction value of the associated components for each fault type in the fault type list to generate fault anomaly early warning information and perform fault anomaly early warning.

[0010] Specifically, the steps for determining whether to activate the barrier gate mechanism include:

[0011] When a passage trigger signal is received, multimodal recognition data is collected and preprocessed. The multimodal recognition data includes vehicle identification data, passage permission associated data, and channel environment data. The passage trigger signal is a valid signal that can trigger the barrier gate to start the passage verification process.

[0012] Based on the preprocessed multimodal recognition data, authentication features are extracted and feature alignment is performed separately. Multidimensional feature vectors are constructed by concatenating them according to the preset feature dimension order.

[0013] The constructed multi-dimensional feature vectors are matched with the preset permission rule base for permission matching, and the environmental impact level features corresponding to the channel environment data are called to dynamically adjust the weight ratio of each dimension of the multi-dimensional feature vector.

[0014] The environmental impact level characteristics are based on channel environmental data and are quantitatively assessed to reflect the graded characteristics of the degree of interference of the environment with dynamic access control verification.

[0015] Specifically, the steps for determining whether to activate the barrier gate mechanism also include:

[0016] Based on the weight ratio, the multi-dimensional feature vector is compared with the preset constraints in the permission rule base dimension by dimension. If all dimensions of the multi-dimensional feature vector match the preset constraints in the permission rule base, the authentication is deemed successful and the barrier gate operation is started.

[0017] If there are multi-dimensional feature vector dimensions that do not match the preset constraints in the permission rule base, the sum of the weight proportions of the matching multi-dimensional feature vector dimensions is used as the feature matching index value.

[0018] If the feature matching index value is greater than the preset feature matching threshold, the authentication is deemed successful, the barrier gate operation is initiated, and a verification mark is made; otherwise, the authentication is deemed unsuccessful, the barrier gate operation is prohibited, and an abnormality warning is triggered simultaneously.

[0019] Specifically, the steps for performing the barrier gate control operation include:

[0020] When it is determined that the barrier gate operation is to be started, the current operating mode parameters are received. The operating mode parameters include the action response priority, power output constraint range, action trajectory curvature range and execution time base.

[0021] Based on the current operating mode parameters, a set of driving parameters for the railing driving instructions is generated, and then railing driving instructions adapted to the current operating mode are generated.

[0022] The barrier drive command includes motion trajectory parameters and power output parameters. The motion trajectory parameters include the upper limit of the lifting angle, the rate change gradient, and the trajectory smoothness coefficient. The power output parameters include the motor voltage adjustment range, the real-time torque compensation value, and the start-stop acceleration threshold.

[0023] The barrier drive command is sent to the barrier execution device, which then operates according to the command, and a feedback monitoring mechanism is activated.

[0024] After the barrier completes the lifting action, the channel status data is collected to construct a channel status time series dataset; and the passage features are extracted to determine whether the target object has left the current channel monitoring area where the barrier is located, and to determine whether there are obstacles in the current channel monitoring area where the barrier is located.

[0025] If the target object has left the current barrier gate's monitoring area and there are no obstacles in the monitoring area, then the barrier reset operation is initiated, generating a barrier reset command; the barrier reset operation is then driven to run, and the feedback monitoring mechanism is activated.

[0026] Specifically, the steps for identifying abnormal operating states and locating abnormal operating characteristics include:

[0027] When the feedback monitoring mechanism is activated, the railing operation data is collected at a preset sampling frequency. The railing operation data includes power system operation data, mechanical motion feedback data, sensor perception data, and command execution response data.

[0028] After performing classification preprocessing on the collected railing operation data, a structured operation dataset is constructed according to a single operation cycle; multi-dimensional operation features are extracted from the operation dataset.

[0029] Call the preset normal operation feature library, which contains standard ranges of operation features and feature association rules for each dimension classified by operation mode;

[0030] The extracted multi-dimensional operational features are compared dimension by dimension with the standard intervals of the corresponding operational features in the normal operation feature library. The deviation between the real-time operational features and the center value of the standard intervals of the operational features is calculated, and the matching degree between each combination of real-time operational features and the corresponding feature association rule is calculated.

[0031] Set deviation judgment threshold and matching degree judgment threshold. If the deviation degree of a single dimension operation feature exceeds the deviation judgment threshold, or the matching degree of a combination of multiple dimension operation features is lower than the matching degree judgment threshold, then the current barrier gate is judged to have a suspected abnormal operation state, and the suspected abnormal operation feature is located.

[0032] If the current feedback monitoring mechanism detects a suspected abnormal operation feature, it retrieves the frequency of the operation cycle in which the suspected abnormal operation feature appears within the preset evaluation period from the operation dataset. If the frequency is greater than the preset frequency threshold, it determines that the current barrier gate is in an abnormal operation state and marks the suspected abnormal operation feature as an abnormal operation feature.

[0033] Specifically, the steps for locating the target fault rule and generating a list of fault types include:

[0034] When an abnormal operating state is identified, based on the operating dataset, the abnormal operating features marked within the evaluation period and the abnormal operating feature values ​​of the corresponding operation period are obtained, and the numerical range is scaled to construct a standardized abnormal feature vector. The abnormal feature vector contains key information such as abnormal operating feature identifier, numerical deviation degree, and occurrence time sequence.

[0035] The abnormal feature vectors are classified according to the data sub-modules to which the abnormal operation characteristics belong, and abnormal feature vector subsets are constructed respectively. The data sub-modules include the power system data sub-module, the mechanical motion data sub-module, the sensor data sub-module, and the command response data sub-module.

[0036] For each subset of abnormal feature vectors, fault rules from a preset fault rule library are selected based on the data submodule label to which the subset of abnormal feature vectors belongs, and a fault rule subset for the corresponding data submodule is constructed. The preset fault rule library stores fault rules according to data submodules, and each fault rule includes the fault type, typical abnormal feature combination and parameter judgment threshold under the corresponding data submodule.

[0037] For each subset of abnormal feature vectors, feature matching is performed with the corresponding subset of fault rules. The total matching degree of each subset of abnormal feature vectors and each fault rule in the corresponding subset of fault rules is calculated. Fault rules with a total matching degree higher than a preset matching threshold are marked as candidate fault rules.

[0038] For each candidate fault rule, the associated verification parameters corresponding to the candidate fault rule are retrieved from the running dataset; the associated verification parameters are compared and verified with the preset normal operating range to identify the target fault rule; and the fault types corresponding to the target fault rules are given in order of total matching degree from high to low, generating a fault type list.

[0039] Specifically, the steps for calculating the total matching degree between each subset of abnormal feature vectors and the corresponding subset of fault rules include:

[0040] The feature identifier matching degree is calculated as the proportion of the number of feature identifiers that match the typical abnormal feature combinations of each fault rule in the subset of abnormal feature vectors to the total number of feature identifiers of fault rules.

[0041] For each matched feature identifier, the numerical deviation of the abnormal operation feature in the abnormal feature vector is compared with the parameter judgment threshold preset by the fault rule. Based on the ratio of the deviation magnitude of the numerical deviation to the range of the parameter judgment threshold, a parameter deviation fit score is generated. The average of the parameter deviation fit scores of all matched features is taken as the comprehensive parameter deviation fit score.

[0042] Set the weight ratio of feature identifier matching degree and comprehensive parameter deviation fit degree, and sum the feature identifier matching degree and comprehensive parameter deviation fit degree according to their respective weights to obtain the total matching degree.

[0043] Specifically, the steps for dynamically switching the operating mode of the barrier gate include:

[0044] The receiving frequency of the barrier gate access trigger signal is obtained in real time. The passage density of the barrier gate per unit time is calculated using the sliding window statistical method. The passage density is compared with the preset flow level threshold to obtain the current passage flow level.

[0045] When the fault type or traffic flow level of the barrier gate is updated, the barrier gate's operating mode switching operation is triggered:

[0046] Receive a list of fault types, combine the preset impact priority of each fault type with the total matching degree of each fault type, calculate the comprehensive ranking score of each fault type, sort the list of fault types, and locate the dominant fault type.

[0047] Based on the mapping rules between the dominant fault type and the preset fault mode, and the mapping rules between the current traffic level and the preset traffic mode, the corresponding fault operation mode and traffic operation mode are located respectively; and based on the preset operation mode priority, the target operation mode is selected from the fault operation mode and traffic operation mode.

[0048] Specifically, the steps for generating fault and anomaly warning information and issuing fault and anomaly warnings include:

[0049] Receive the list of fault types, call the preset list of controlled fault types and match each fault type in the list one by one to identify the current controlled fault type of the barrier gate.

[0050] Based on the priority of the impact of the controlled fault type and combined with the current operating mode parameters, temporary control instructions for each controlled fault type are generated sequentially.

[0051] Call the preset fault component association mapping table to locate the associated components for each fault type, retrieve the cumulative runtime data of each associated component, and generate the runtime sequence of each associated component.

[0052] The total matching degree of each fault type is converted into a fault impact coefficient, which is used to quantify the additional attenuation of the lifespan of the associated components due to the current fault type.

[0053] For each associated component, the appropriate time-series decay algorithm model is invoked. The design life baseline value of the associated component is used as the initial value. The normal decay amount under fault-free conditions is calculated based on the runtime sequence, and the additional decay amount of the current fault type is added to obtain the total cumulative loss of the associated component. Then, the remaining life prediction value of the associated component for each fault type is obtained.

[0054] Based on the priority of the impact of the controlled fault types and the predicted remaining life of the components associated with each fault type, early warning levels are divided, and fault anomaly early warning information is integrated to conduct fault anomaly early warning.

[0055] The intelligent control method for barrier gates based on operational status feedback includes:

[0056] Step S1: Upon receiving the passage trigger signal, collect multimodal recognition data, perform authentication feature extraction and permission rule base matching on the multimodal recognition data, perform dynamic permission verification, and determine whether to start the barrier gate operation;

[0057] Step S2: If the barrier gate operation is started, based on the current operating mode of the barrier gate, the barrier gate control operation is executed, barrier drive commands and barrier reset commands are generated, the barrier operation is controlled, and the feedback monitoring mechanism is started;

[0058] Step S3: When the feedback monitoring mechanism is activated, the railing operation data is collected, the operation dataset is constructed according to the operation cycle, and the abnormal operation status is identified and the abnormal operation characteristics are located by combining the normal operation feature library of the current operation mode; the abnormal operation characteristics are then classified and matched with the fault rule library to locate the target fault rule and generate a fault type list.

[0059] Step S4: Update the current passage flow level of the barrier gate according to the receiving frequency of the passage trigger signal, and locate the dominant fault type from the fault type list, so as to dynamically switch the operating mode of the barrier gate in combination with the passage flow level;

[0060] Step S5: Identify the controlled fault types in the fault type list, generate temporary control commands for each controlled fault type in sequence; and calculate the remaining life prediction value of the associated components for each fault type in the fault type list to generate fault anomaly warning information and perform fault anomaly warning.

[0061] The beneficial effects of this invention are:

[0062] This application achieves intelligent control of barrier gates through multi-module collaboration, significantly improving the accuracy of traffic control, operational adaptability, and effectiveness of fault handling: Multimodal recognition data collection combined with a permission verification method that dynamically adjusts feature weights based on environmental impact levels effectively avoids the problem of single-module recognition being affected by environmental interference, greatly reducing the probability of falsely intercepting legitimate targets or allowing unauthorized passage, thus ensuring traffic order and safety; Dynamic switching of operating modes based on traffic flow levels and dominant fault types flexibly adapts to response needs under different traffic scenarios, improving response efficiency during peak hours to alleviate congestion, and optimizing parameters during off-peak hours to reduce energy consumption, balancing traffic efficiency and operational economy; Through full-dimensional operational data collection, multi-feature comparison to identify anomalies, and classification matching to locate faults, rapid and accurate fault detection is achieved, avoiding the lag of manual inspections or single threshold judgments; Temporary control commands are generated for control faults to promptly suppress fault deterioration, and graded early warning is conducted based on the remaining lifespan prediction of related components, enabling advance planning of maintenance plans, reducing downtime caused by sudden failures of key components, lowering operation and maintenance costs, and ultimately achieving efficient, stable, and safe operation of barrier gates in different scenarios, improving the overall traffic experience and equipment management level. Attached Figure Description

[0063] Figure 1 This is a schematic diagram of the intelligent control system for barrier gates based on operational status feedback according to the present invention.

[0064] Figure 2 This is a flowchart illustrating the process of determining whether to activate the barrier gate mechanism according to the present invention.

[0065] Figure 3 This is a flowchart for identifying abnormal operating states in this invention;

[0066] Figure 4 This is a flowchart illustrating the dynamic switching operation mode of the barrier gate mechanism according to the present invention.

[0067] Figure 5 This is a flowchart of the intelligent control method for barrier gates based on operational status feedback according to the present invention. Detailed Implementation

[0068] Please see Figure 1 This embodiment introduces an intelligent control system for a barrier gate based on operational status feedback, including a status perception module, an analysis and diagnosis module, a feedback control module, and an early warning interaction module.

[0069] When the status perception module receives a passage trigger signal, it simultaneously collects multimodal recognition data, including vehicle identification data, passage permission association data, and channel environment data. Using a feature-level fusion algorithm, it extracts authentication features and matches them against the permission rule base to perform dynamic permission verification and determine whether to activate the barrier gate operation. If so, based on the current barrier gate operating mode, it executes barrier gate control operations, generating barrier drive and reset commands. Specifically, it generates barrier drive commands adapted to the current operating mode, combining channel environment data. These commands include motion trajectory parameters and power output parameters, used to drive the barrier execution device to raise the barrier. While the barrier remains in the open state, it collects channel status data through multiple sensors, including target obstacle status data and channel obstacle monitoring data. Using a time-series analysis algorithm, it performs real-time analysis of the channel status data to determine if the target has completely passed through the channel. If passage is confirmed, a matching barrier reset command is generated, driving the barrier execution device to orderly lower the barrier, achieving closed-loop execution of the passage process.

[0070] In this embodiment, the state perception module synchronously collects multimodal recognition data and uses a feature-level fusion algorithm for correlation processing, breaking through the limitations of traditional single recognition dimensions. This enables dynamic verification of access permissions, significantly improving the accuracy and comprehensiveness of permission authentication and effectively avoiding problems such as misjudgment and interception or unauthorized release caused by blind spots or data fragmentation. By combining the current operating mode, channel environment data, and historical execution records to generate adaptive barrier drive commands and barrier reset commands, it overcomes the rigidity of traditional fixed command execution, achieving precise adaptation of barrier actions to the operating scenario and improving the stability and targeting of barrier operation. By using multiple sensors to collaboratively collect channel state data and analyzing it in real time through time-series analysis algorithms, it comprehensively captures the status of target objects and obstacle information within the channel, ensuring timely and accurate judgment of whether the target has completely passed through the channel. This avoids the safety risks caused by barrier mis-falling due to incomplete channel state monitoring and ensures the closed-loop and efficient execution of the release process, thereby improving the overall intelligence, safety, reliability, and execution efficiency of the barrier gate access control.

[0071] Please see Figure 2 Preferably, the specific steps for determining whether to start the barrier gate operation include:

[0072] Upon receiving a passage trigger signal, the system immediately initiates multi-source sensing and acquisition operations to collect multimodal recognition data and perform data preprocessing. This multimodal recognition data includes vehicle identification data, passage permission associated data, and channel environment data. Specifically, the passage trigger signal refers to a valid signal that can trigger the barrier gate to initiate the passage verification process, including passage request signals initiated by the vehicle owner and target presence signals passively detected by the equipment. Passage request signals initiated by the vehicle owner include card swiping signals, QR code scanning signals, and ETC sensing signals; target presence signals passively detected by the equipment include ground loop detection signals and visual sensor target capture signals. Data preprocessing includes: performing character normalization and tilt correction preprocessing on vehicle identification data; performing format standardization and associated field extraction preprocessing on passage permission associated data; performing outlier removal and data smoothing preprocessing on channel environment data; filtering electromagnetic interference signals during the acquisition process using an adaptive filtering algorithm; and removing redundant and invalid data using a feature filtering algorithm.

[0073] Based on the preprocessed multimodal recognition data, a unified format authentication feature is extracted using a multi-dimensional feature extraction and alignment strategy. Examples include: for vehicle identification data, character structure features are extracted using contour extraction algorithms and morphological processing; character texture features are extracted using gray-level co-occurrence matrix; and character layout features are extracted using character position coordinate mapping. For access permission association data, attribute information such as permission validity period, access area, and vehicle type is transformed into permission attribute features using a field encoding algorithm; and device binding relationship features are extracted using an association degree calculation algorithm. For channel environment data, threshold segmentation algorithms are used to extract features related to light intensity and weather conditions. Environmental factors such as conditions are transformed into environmental impact level features, which are then extracted through gradient calculation. Feature alignment processing is performed on each of the extracted authentication features, mapping different dimensions of authentication features to the same feature space via coordinate mapping. A Min-Max normalization algorithm is used to uniformly scale all authentication features to the same numerical range, ensuring that the authentication features extracted from multimodal data corresponding to different trigger signals have structural consistency and comparability. Based on the aligned and normalized features, multi-dimensional feature vectors are constructed by concatenating them according to a preset feature dimension order, providing a unified format for feature input for subsequent dynamic verification of access permissions.

[0074] The constructed multi-dimensional feature vectors are matched with a preset permission rule base for permission matching. The permission rule base includes rules for access permission validity, access area matching, vehicle type adaptation, and device binding legality. During permission matching, the weight ratio of each dimension of the multi-dimensional feature vector is dynamically adjusted by calling the environmental impact level features corresponding to the channel environment data. The environmental impact level features are based on channel environment data, quantitatively evaluated, and reflect the degree of interference of the environment on dynamic permission verification. They are obtained through a combination of multi-parameter fusion evaluation and threshold interval division. Specifically, key environmental parameters such as light intensity, rain / snow level, visibility, and road surface interference density in the channel environment data are first quantified to obtain quantified values ​​for each parameter. The analytic hierarchy process (AHP) is used to determine the impact weight of each environmental parameter on identity recognition and permission verification. The quantified values ​​of each parameter are multiplied by their corresponding weights and summed to obtain the comprehensive environmental impact assessment value. The comprehensive environmental impact assessment value is compared with the preset level division thresholds; falling into different threshold intervals corresponds to different environmental impact level features. The environmental impact level features include... The environmental impact level is categorized into low, medium, and high impact levels. For example, when the environmental impact level is determined to be low, the verification weight of vehicle identification features and permission association features is increased; when the environmental impact level is determined to be high, the basic verification weight is retained. Based on the weight percentage, the multi-dimensional feature vector is compared dimension-by-dimensionally with the preset constraints in the permission rule base. If all dimensions of the multi-dimensional feature vector match the preset constraints in the permission rule base, authentication is considered successful, and the barrier gate operation is initiated. If any dimension of the multi-dimensional feature vector matches the preset constraints in the permission rule base, the authentication is considered successful, and the barrier gate operation is initiated. If the constraints do not match, the sum of the weight percentages of the matching multi-dimensional feature vector dimensions is used as the feature matching index value. If the feature matching index value is greater than the preset feature matching threshold, the authentication is deemed successful, the barrier gate operation is initiated, and a verification mark is made. This is used to trace the feature matching details, unmatched dimensions, and weight allocation of this verification, providing accurate traceability data support for iterative optimization of the dynamic weight adjustment strategy and the constraint threshold of the permission rule base. The preset feature matching threshold is a critical value used to determine whether the access permission is qualified in scenarios where some features do not match. It is a quantitative standard for measuring the degree of matching between the multi-dimensional feature vector and the permission rule base. It is determined through statistical analysis and iterative verification, taking into account the core degree of each dimension feature in the permission rule base, the distribution range of feature matching index values ​​in normal access scenarios in historical verification data, and the balance between access security requirements and access efficiency requirements. Otherwise, the authentication is deemed unsuccessful, the barrier gate operation is prohibited, and an abnormal prompt warning is triggered simultaneously to provide feedback to the access target on the core reason for the authentication failure. At the same time, abnormal verification data and feature matching logs are uploaded to the operation and maintenance system, providing clear data guidance for operation and maintenance personnel to troubleshoot and identify faults and correct permission information, thus completing the access permission verification process.

[0075] Preferably, the specific steps for performing the barrier gate control operation include:

[0076] When it is determined that the barrier gate operation is to be started, the current operating mode parameters issued by the feedback control module are received through the preset communication protocol. The operating mode parameters include the action response priority, power output constraint range, action trajectory curvature range and execution time base, which clarifies the core execution standard of the barrier operation in the current scenario.

[0077] Based on the current operating mode parameters, the reference value of the lifting angle of the guardrail is determined according to the action response priority and execution time reference. The reference value of the motor drive voltage and the basic adjustment value of the torque are defined in combination with the power output constraint range. The initial curve of the lifting rate is fitted with reference to the curvature range of the action trajectory to generate the drive parameter set of the guardrail drive command. The drive parameter set includes the reference value of the lifting angle of the guardrail, the initial curve of the lifting rate, the reference value of the motor drive voltage and the basic adjustment value of the torque.

[0078] Based on the drive parameter set, the upper limit of the lifting angle and the rate change gradient are set according to the trajectory smoothness requirements. The trajectory smoothness coefficient is calculated by the trajectory smoothing algorithm. The motor voltage adjustment range is determined by reserving the load fluctuation adaptation space, and the real-time torque compensation value is generated by superimposing the torque base adjustment value with dynamic load compensation logic. The start-stop acceleration threshold is set by matching the rate change gradient. The barrier drive command adapted to the current operating mode is generated. The barrier drive command includes motion trajectory parameters and power output parameters. The motion trajectory parameters include the upper limit of the lifting angle, the rate change gradient, and the trajectory smoothness coefficient. The power output parameters include the motor voltage adjustment range, the real-time torque compensation value, and the start-stop acceleration threshold.

[0079] The barrier drive command is sent to the barrier execution device, which then operates according to the command to lift the barrier; a feedback monitoring mechanism is activated at the same time.

[0080] After the barrier completes the lifting action, the channel status data is collected through the collaborative monitoring mechanism of infrared array sensors and lidar, including the real-time position coordinates, movement speed vector, and contour size parameters of the target object, as well as the distance measurement, spatial size data, and movement direction vector of obstacles in the channel; the collected channel status data is structured and integrated in the order of timestamps to construct a channel status time series dataset;

[0081] Traffic features are extracted from the time-series dataset of channel status. These features include the positional change trend of the target object, velocity decay features, contour departure features, and spatial occupancy features of obstacles within the channel.

[0082] Based on traffic characteristics, the system compares the real-time location coordinates of the target object with the boundary coordinates of the preset channel monitoring area. It also verifies whether the vector magnitude of the target object's moving speed is lower than a preset speed threshold within multiple consecutive sampling periods. Combined with whether the infrared array sensor fails to capture the complete outline of the target object continuously, it determines whether the target object has left the monitoring area of ​​the current barrier gate. Furthermore, by matching the distance measurement values ​​of obstacles in the channel collected by the lidar with the preset monitoring distance range, it checks whether the spatial size data of the obstacles exceeds the preset obstacle size threshold. It confirms that no new obstacles have been added in the channel within multiple consecutive sampling periods and that the original obstacles have been completely moved out of the monitoring range, thus determining whether there are obstacles in the monitoring area of ​​the current barrier gate.

[0083] If the target object has left the current barrier gate's monitoring area and there are no obstacles in the monitoring area, then the barrier gate reset operation is initiated, generating a barrier gate reset command.

[0084] The channel status time-series dataset contains the spatiotemporal characteristics of the target and obstacles at different times. It is used to analyze the progress of the target in real time using time-series analysis algorithms, determine whether the target has completely left the channel monitoring area, and continuously monitor for new obstacles appearing within the channel. Based on the motion trajectory curvature range in the operation mode parameters, a falling angle change curve is fitted. Combined with the power output constraint range, a power output attenuation coefficient is set to generate a barrier reset command. The barrier reset command includes the falling angle change curve, the power output attenuation coefficient, and the falling endpoint positioning parameters. The barrier reset command is then sent to the barrier execution device, driving the barrier reset operation and causing the barrier to fall orderly along a preset trajectory.

[0085] The analysis and diagnosis module collects barrier operation data when the feedback monitoring mechanism is activated. This data includes power system operation data, mechanical motion feedback data, sensor perception data, and command execution response data. The collected data is structured and organized according to a single operation cycle to construct an operation dataset containing operation sequence, operating parameters, and environmental correlation information. This dataset comprehensively reflects the full-dimensional operating status of the barrier mechanism for each operation. Based on this dataset, data features are extracted and analyzed. By comparing the deviations between the normal operation feature library and real-time data features, abnormal operating states are identified, including barrier jamming, motor overload, and abnormal sensor signals. Simultaneously, by combining the correlations of abnormal features with a fault rule library, the fault type is located, providing accurate fault basis for subsequent feedback adjustment and early warning. This overcomes the limitations of traditional single-threshold judgments and improves the accuracy of anomaly identification and fault location.

[0086] In this embodiment, by collecting comprehensive barrier operation data, including power system operation data and mechanical motion feedback data, and organizing the data into a structured dataset containing operation sequence, operating parameters, and environmental correlation information according to a single operation cycle, a comprehensive capture of the barrier machine's operating status for each operation is achieved, avoiding the omissions caused by the one-sided nature of traditional data collection. By extracting features from the operating dataset and comparing them with a normal operation feature library, and associating abnormal features with a fault rule library, the limitations of traditional single threshold judgment are overcome. This not only accurately identifies various abnormal operating states such as barrier jamming, motor overload, and abnormal sensor signals, but also accurately locates the corresponding fault types, significantly improving the comprehensiveness of abnormal identification and the accuracy of fault location. This provides reliable fault basis for the dynamic adjustment of the feedback control module and the accurate early warning of the early warning interaction module, effectively shortening fault troubleshooting time, reducing maintenance costs, and ensuring the long-term stable and efficient operation of the barrier machine.

[0087] Please see Figure 3 Preferably, the specific steps for identifying abnormal operating states include:

[0088] When the feedback monitoring mechanism is activated, the railing operation data is collected in real time at a preset sampling frequency. The railing operation data includes power system operation data, mechanical motion feedback data, sensor perception data, and command execution response data. Among them, the power system operation data includes motor voltage, operating current, output torque, and core component temperature data; the mechanical motion feedback data includes railing lifting and lowering angles, action execution duration, trajectory offset, and joint force data; the sensor perception data includes real-time output signals and data transmission status of angular displacement sensors, Hall sensors, infrared array sensors, and lidar; and the command execution response data includes command reception timestamps, execution start delays, and deviation data between actual action parameter values ​​and preset command values.

[0089] The collected railing operation data undergoes classification preprocessing, while the power system operation data undergoes outlier removal and data smoothing preprocessing. Extreme outliers are identified and removed using the 3σ criterion, and data smoothing is achieved using the moving average method. Mechanical motion feedback data undergoes format standardization and time-series alignment preprocessing, converting feedback data from different sensors into a pre-defined data format and aligning the time sequence based on the command issuance timestamp. Sensor sensing data undergoes signal noise reduction and invalid data filtering preprocessing, using a Kalman filter algorithm to reduce signal noise and remove invalid data with interrupted data transmission or incorrect formats. Command execution response data undergoes delay time calibration and response status encoding preprocessing, calibrating the delay time based on the system clock and converting the response status into a standardized code.

[0090] Based on the pre-processed barrier operation data, a structured operation dataset is constructed according to a single operation cycle: a complete operation cycle consists of barrier drive command issuance, barrier raising, passage monitoring, barrier resetting, and completion of the action. Within each operation cycle, various types of pre-processed barrier operation data are integrated in time stamp order to form a structured entry containing operation cycle identifiers, time series axes, and data modules. Among them, the data modules are divided into power system data sub-modules, mechanical motion data sub-modules, sensor data sub-modules, and command response data sub-modules, and each sub-module stores the corresponding type of time series data and statistical parameters.

[0091] Multi-dimensional operational features are extracted from the constructed operational dataset. For the power system data submodule, a sliding window algorithm is used to extract time-series variation features such as parameter mean, fluctuation amplitude, peak frequency, and rate of change. For the mechanical motion data submodule, a statistical analysis algorithm is used to extract motion accuracy features such as angle deviation, motion completion time deviation, trajectory offset peak, and joint force distribution. For the sensor data submodule, a gradient calculation algorithm is used to extract signal stability features such as signal intensity fluctuation range, average data transmission delay, effective data ratio, and number of signal mutations. For the command response data submodule, a correlation analysis algorithm is used to extract response matching features such as command reception delay, execution parameter deviation, state feedback consistency, and command retransmission count.

[0092] Call the preset normal operation feature library, which is built based on a large amount of normal operation data under different operation modes. It includes standard intervals of operation features in each dimension classified by operation mode and feature association rules. Each standard interval of operation features corresponds to the standard range of parameters under normal operation. The feature association rules define the mutual constraint relationship of operation features in each dimension under normal operation.

[0093] The extracted multi-dimensional operational features are compared dimension by dimension with the standard intervals of the corresponding operational features in the normal operation feature library. The deviation between the real-time operational features and the center value of the standard intervals of operational features is calculated, and the matching degree between each combination of real-time operational features and the feature association rules is calculated.

[0094] Set deviation judgment threshold and matching degree judgment threshold. The deviation judgment threshold and matching degree judgment threshold are determined by statistically analyzing the characteristic fluctuation range of normal operation data. If the deviation degree of a single dimension operation feature exceeds the deviation judgment threshold, or the matching degree of a combination of multiple dimension operation features is lower than the matching degree judgment threshold, then the current barrier gate is judged to have a suspected abnormal operation state, and the suspected abnormal operation feature is located.

[0095] If the current feedback monitoring mechanism detects suspected abnormal operation characteristics, it retrieves the frequency of operation cycles in which the suspected abnormal operation characteristics appear within the preset evaluation period from the operation dataset. If the frequency is greater than the preset frequency threshold, it determines that the current barrier gate is in an abnormal operation state and marks the suspected abnormal operation characteristic as an abnormal operation characteristic. The preset evaluation period is set in conjunction with the average daily operation frequency of the barrier gate, and is usually 1 to 2 hours, covering 200 to 500 operation cycles. The preset frequency threshold is determined based on the statistical analysis of historical normal operation data.

[0096] Preferably, the specific steps for locating the fault type include:

[0097] When an abnormal operating state is identified, based on the operating dataset, the marked abnormal operating features within the evaluation period and the corresponding abnormal operating feature values ​​for the corresponding operating period are obtained. The abnormal operating feature values ​​are then uniformly scaled to the same numerical range using a Min-Max normalization algorithm to construct a standardized abnormal feature vector. This vector contains key information such as the abnormal operating feature identifier, the degree of numerical deviation, and the occurrence sequence. The abnormal operating feature identifier is a unique label used to distinguish different types of abnormal operating features. The degree of numerical deviation refers to the deviation of the real-time monitoring value of the abnormal operating feature from the corresponding standard range in the normal operating feature database; it is a core indicator for quantifying the severity of the abnormality. This is obtained by calculating the difference between the real-time value of the abnormal operating feature and the center value of the standard range for normal operating features, and then normalizing it by combining the range of the standard range. The occurrence sequence refers to the time information of the first occurrence and subsequent recurrence of the abnormal operating feature within a preset evaluation period, including a specific timestamp and the corresponding operating period number, used to trace the temporal pattern of the abnormality.

[0098] The abnormal feature vectors are classified according to the data sub-modules to which the abnormal operation characteristics belong. The data sub-modules include the power system data sub-module, the mechanical motion data sub-module, the sensor data sub-module, and the command response data sub-module. Abnormal feature vector subsets are constructed for each of these sub-modules, namely: power system abnormal feature vector subset, mechanical motion abnormal feature vector subset, sensor abnormal feature vector subset, and command response abnormal feature vector subset. Each abnormal feature vector subset contains only the abnormal feature information of the corresponding data sub-module, ensuring that abnormal features of similar system modules are analyzed in a concentrated manner.

[0099] For each subset of abnormal feature vectors, fault rules from a pre-defined fault rule library are selected based on the data submodule label to which the subset of abnormal feature vectors belongs, thus constructing a fault rule subset for the corresponding data submodule. The pre-defined fault rule library stores fault rules categorized by data submodule, and each fault rule includes the fault type, typical abnormal feature combinations, and parameter judgment thresholds for the corresponding data submodule. For example, the abnormal feature vector subset of the power system corresponds to all rules under the power system category in the fault rule library, such as motor overload and voltage instability fault rules, forming a power system fault rule subset.

[0100] For each subset of abnormal feature vectors, feature matching is performed with the corresponding subset of fault rules. The total matching degree of each subset of abnormal feature vectors with each fault rule in the corresponding subset of fault rules is calculated, including: the proportion of the number of feature identifiers in the subset of abnormal feature vectors that match the typical abnormal feature combinations of fault rules to the total number of feature identifiers of fault rules, which is used as the feature identifier matching degree.

[0101] For each matched feature identifier, the numerical deviation of the abnormal operation feature in the abnormal feature vector is compared with the preset parameter judgment threshold of the fault rule. If the numerical deviation is within the parameter judgment threshold range, the parameter deviation fit score of the abnormal operation feature is set to full score; otherwise, the parameter deviation fit score is generated based on the ratio of the magnitude of the numerical deviation exceeding the corresponding parameter judgment threshold range. For example, if the numerical deviation of the abnormal operation feature is within the corresponding parameter judgment threshold range, the parameter deviation fit score is set to full score; otherwise, the ratio of the magnitude of the numerical deviation of the abnormal operation feature to the preset allowable deviation upper limit of the corresponding parameter judgment threshold range is calculated, and the score decreases linearly as the magnitude ratio increases, thus generating the parameter deviation fit score. The average of the parameter deviation fit scores of all matched features is taken as the comprehensive parameter deviation fit score.

[0102] Based on the importance of each feature dimension in the fault rule for fault determination, the weight ratios of feature identifier matching degree and comprehensive parameter deviation fit are set; the total matching degree is the result of weighted calculation of the scores of each dimension according to their corresponding weights, and the score range is between the lowest benchmark value and the full score, which is used to quantify the degree of fit between the abnormal feature vector subset and the target fault rule.

[0103] For each subset of abnormal feature vectors, fault rules with a total matching degree higher than a preset matching threshold are marked as candidate fault rules. The preset matching threshold is set based on the importance of the core features of the fault rules and the effectiveness of historical matching, ensuring that the selected candidate fault rules have a high degree of fit with the subset of abnormal feature vectors.

[0104] For each candidate fault rule, the associated verification parameters are retrieved from the runtime dataset. These parameters are key auxiliary parameters that inevitably fluctuate when the fault type occurs. For example, a motor overload fault rule is associated with motor winding temperature and cooling system operating status parameters; a railing jamming fault rule is associated with peak force on mechanical joints and transmission component speed fluctuation parameters. The associated verification parameters are compared with the preset normal operating range. If the associated verification parameters are not within the normal operating range, the corresponding candidate fault rule is marked as the target fault rule, and only candidate fault rules with abnormal fluctuations in associated verification parameters are retained. For the target fault rules, they are sorted from high to low total matching degree, and their corresponding fault types are given in order, generating a fault type list.

[0105] The feedback control module acquires the reception frequency of the passage trigger signal collected by the status perception module, calculates the passage density using a sliding window statistical method, and determines the current passage flow level based on a preset threshold. Simultaneously, it receives the fault type list and associated fault impact range output by the analysis and diagnosis module, dynamically switching the barrier gate's operating mode. It presets adaptation modes for different passage flow levels and fault-tolerant operating modes for various faults. Each mode is configured with differentiated action response priorities, power output constraints, action trajectory curvature intervals, execution time bases, and safety thresholds: the high-flow mode emphasizes rapid response to improve passage efficiency, the low-flow mode emphasizes stable operation to reduce equipment wear, and the fault-tolerant mode optimizes parameters for specific fault types to prevent fault escalation and maintain basic passage functionality. The module sends the switched operating mode parameters to the status perception module in real time, simultaneously and dynamically fine-tuning the power output and action trajectory parameters in the barrier drive and reset commands based on the fault type, ensuring that the barrier gate can balance passage efficiency, equipment safety, and operational stability under different flow scenarios and fault states.

[0106] In this embodiment, the feedback control module overcomes the rigid limitations of the traditional fixed operation mode of barrier gates by constructing a dual-dimensional dynamic adjustment mechanism based on flow level and fault type. It precisely switches between modes based on real-time traffic flow, significantly reducing single-pass time and effectively alleviating congestion in high-flow scenarios by improving response speed. In low-flow scenarios, it reduces mechanical impact and energy consumption by optimizing stability parameters, extending equipment lifespan. Simultaneously, it dynamically activates a fault-tolerant operation mode for different fault types, preventing fault escalation through parameter fine-tuning while ensuring basic passage functionality, solving the problems of easy downtime or excessive wear under traditional fault conditions. This on-demand, fault-tolerant closed-loop adjustment capability enables the barrier gate to achieve an optimal balance of passage efficiency, equipment safety, and operational stability under complex flow fluctuations and equipment malfunctions, significantly improving the system's intelligent control level and practical application adaptability.

[0107] Please see Figure 4Preferably, the specific steps for dynamically switching the operating mode of the barrier gate include:

[0108] The system acquires the receiving frequency of the barrier gate's passage trigger signal in real time, calculates the passage density per unit time using a sliding window statistical method, and compares the passage density with a preset traffic level threshold to obtain the current traffic level. This provides a basis for scenario adaptation in mode matching. The traffic level threshold refers to the critical numerical standard of passage density per unit time used to classify different traffic levels such as high, medium, and low. It is the core basis for determining the traffic intensity of the current passage scenario and matching the corresponding operating mode. The threshold is set in combination with the passage patterns of the barrier gate deployment scenario, the statistical analysis results of historical passage data, the equipment's operating performance limits, and the need to balance passage efficiency and equipment wear.

[0109] When the fault type or traffic flow level of the barrier gate is updated, a mode switching operation is triggered. The system receives a fault type list from the analysis and diagnosis module, and sorts the list using a weighted comprehensive ranking method. The priority of the fault type is the primary weight, and the total matching degree is a secondary weight. A comprehensive ranking score is calculated for each fault type, and the fault types are ranked from highest to lowest score. The fault type ranked first is the current dominant fault type, ensuring that mode switching prioritizes responding to the faults most critical to system operation. The preset priority of the fault type refers to a pre-defined grading standard that measures the critical impact of different fault types on the safe operation of the barrier gate system, the maintenance of traffic efficiency, and the loss of core equipment performance. This clarifies the threat level ranking of faults to the overall system operation. The system analyzes the occurrence mechanism of various faults and their impact weights on system safety risks, traffic interruption probability, and equipment lifespan degradation, combining historical fault handling case data with the safety priority requirements of actual application scenarios.

[0110] Based on the mapping rules between the dominant fault type and the preset fault mode, and the mapping rules between the current traffic level and the preset traffic mode, the corresponding fault operation mode and traffic operation mode are located. The preset fault mode mapping rule specifies the dedicated fault-tolerant operation mode corresponding to each type of fault, and the preset traffic mode mapping rule specifies the dedicated adaptive operation mode corresponding to each traffic level. Both mapping rules are built based on the system operation logic and actual application scenarios to ensure the accuracy of mode location.

[0111] Based on the preset operating mode priority, the operating mode with higher priority between the fault operating mode and the flow operating mode is selected as the target operating mode. The operating mode priority is determined according to the strength of the constraints of the operating mode parameters. The parameter constraints are reflected in the strength of the restriction on the operation of the equipment, including the strictness of the upper limit of power output, the conservative range of the motion trajectory, and the stringency level of the safety threshold. The stronger the constraint, the higher the priority of the mode.

[0112] Extract complete parameters of the target operating mode from the preset operating mode parameter library, including action response priority, power output constraint range, action trajectory curvature range, execution time base and safety threshold, to ensure that the extracted parameters fully cover the operating requirements of the target mode and are compatible with the dominant fault type and the current flow level.

[0113] The extracted target operating mode parameters are sent to the status perception module in real time through a preset communication protocol, and the instruction optimization mechanism is triggered simultaneously. Based on the adaptation requirements of the dominant fault type and the current flow level, the power output parameters and motion trajectory parameters in the barrier drive instruction and barrier reset instruction are dynamically fine-tuned, so that the instruction and the target operating mode are accurately matched, and the mode parameters are effectively implemented.

[0114] The early warning interaction module receives the fault type list output by the analysis and diagnosis module and synchronously obtains the current operating mode parameters issued by the feedback control module. It matches the fault type list with a preset list of controlled fault types to identify whether there are any critical controlled fault types. If so, it generates targeted temporary control commands by combining core information such as power output constraints and motion trajectory parameters in the current operating mode. It suppresses fault deterioration and maintains basic passage functions by adjusting the upper limit of power output and optimizing the action execution logic. At the same time, it calls the corresponding component remaining life prediction model according to the system module category to which the fault belongs. It integrates information such as the degree of fault impact, the cumulative running time of the equipment, and historical wear data to quantitatively predict the remaining life of the related core components. Finally, it classifies the early warning level according to the severity of the fault and the remaining life threshold, and issues anomaly warnings. It pushes fault details, remaining life prediction results, and maintenance suggestions to operation and maintenance personnel through multiple channels, and provides temporary operation adjustment prompts to the passage targets to ensure the safety of equipment operation and the timeliness of problem handling.

[0115] Preferably, the specific steps for issuing fault and anomaly warnings include:

[0116] Receive the list of fault types, call the preset list of controlled fault types, and match each fault type in the list of fault types with the list of controlled fault types one by one to identify the current controlled fault type of the barrier gate. The list of controlled fault types includes key fault categories that directly affect safe operation, such as motor overload, barrier jamming, and sensor signal failure.

[0117] Based on the priority of the impact of the controlled fault type and combined with the current operating mode parameters, temporary control commands for each controlled fault type are generated sequentially. For example, for power system controlled faults, the motor voltage regulation range and the upper limit of the real-time torque compensation value are adjusted; for mechanical motion controlled faults, the motion trajectory curvature range and start-stop acceleration threshold are optimized; for sensor controlled faults, the cross-validation logic of multi-sensor data is strengthened to ensure that the temporary control commands can both suppress the deterioration of the fault and adapt to the current operating scenario.

[0118] Extract the fault type identifier from the fault type list, call the preset fault component association mapping table. The fault component association mapping table is constructed based on the equipment structure and fault mechanism, and clearly defines the one-to-one correspondence between motor overload and drive motor, railing jamming and transmission gear set, and sensor signal abnormality and corresponding sensor, etc., to locate the associated components directly affected by each fault type and ensure the accuracy of component matching.

[0119] The cumulative runtime data of each related component is retrieved, including total runtime, runtime percentage under different load levels, and runtime after historical fault repair. Through timestamp calibration and format standardization, a runtime sequence of each related component is formed, providing time-dimensional basic data for lifespan degradation analysis.

[0120] The total matching degree of each fault type is converted into a fault impact coefficient. The higher the total matching degree, the greater the fault impact coefficient. If the total matching degree is full, the coefficient takes the upper limit value. If it is lower than the preset threshold, the coefficient is reduced proportionally. This is used to quantify the additional attenuation of the life of related components by the current fault type, and to ensure the differentiated manifestation of the fault impact.

[0121] For different types of related components, the corresponding time-series attenuation algorithm model is called. For example, the mechanical transmission component adopts the fatigue strength attenuation model, which takes into account the cumulative running time, the proportion of heavy load cycles, and the fault influence coefficient, and calculates the normal attenuation and additional attenuation through the Miner fatigue accumulation rule; the electric drive component adopts the thermal aging attenuation model, which takes into account the cumulative running time, the average operating current, and the temperature fluctuation data of the core component, and calculates the degree of insulation aging by combining the fault influence coefficient; the sensing and detection component adopts the signal stability attenuation model, which incorporates the cumulative running time, the frequency of abnormal data transmission, and the fault influence coefficient, and calculates the performance loss through the signal attenuation rate formula.

[0122] In the time-series decay algorithm, the design life baseline value of the associated components is used as the initial value. The normal decay amount under fault-free conditions is calculated based on the runtime sequence, and the additional decay amount of the current fault type is added to obtain the total cumulative loss of the associated components. The total cumulative loss is then subtracted from the design life baseline value to obtain the predicted remaining life value of the associated components for each fault type.

[0123] Based on the priority of the impact of the controlled fault types and the predicted remaining lifespan of the associated components for each fault type, early warning levels are classified, and fault anomaly early warning information is integrated: the content is sorted according to the early warning level, including the controlled fault type, the name and predicted remaining lifespan of the associated components, details of temporary control instructions, and targeted handling suggestions, such as stopping the machine to check the winding insulation for motor overload, and adding special lubricant for gear jamming, forming a structured early warning information package for fault anomaly early warning; the complete early warning information package is sent to the operation and maintenance management platform, triggering pop-up prompts and work order generation; the equipment locally activates the audible and visual early warning device; a concise prompt is pushed to the channel entrance display screen, such as temporary equipment maintenance, which may delay passage, ensuring that operation and maintenance personnel have detailed handling basis and that the passage targets are aware of the operating status.

[0124] Please see Figure 5 This embodiment introduces an intelligent control method for barrier gates based on operational status feedback, including:

[0125] Step S1: Receive the passage trigger signal, synchronously collect multimodal recognition data composed of vehicle identification data, passage permission associated data and channel environment data, preprocess the multimodal recognition data, extract authentication features in a unified format through feature-level fusion algorithm, match the authentication features with the permission rule base and dynamically adjust the feature weight ratio to complete dynamic permission verification and determine whether to start the barrier gate operation.

[0126] Step S2: If the barrier gate operation is started, the current operating mode parameters issued by the feedback control module are received. Combined with the channel environment data, a barrier drive command containing motion trajectory parameters and power output parameters is generated to drive the barrier execution device to complete the lifting action and start the feedback monitoring mechanism. During the barrier release period, channel status data is collected through multi-sensor collaboration and a time series dataset is constructed. Based on the time series analysis algorithm, the status of the target passage and the channel obstacle situation are analyzed. When the target has completely passed and there are no obstacles in the channel, a barrier reset command is generated to drive the barrier execution device to fall in an orderly manner.

[0127] Step S3: After the feedback monitoring mechanism is activated, power system operation data, mechanical motion feedback data, sensor perception data, and command execution response data are collected. The data is classified and preprocessed, and an operation dataset is constructed according to a single operation cycle. Multi-dimensional operation features are extracted from the dataset and compared with the normal operation feature library. Abnormal operation status is identified by deviation and matching degree. Then, abnormal features are combined with the fault rule library for feature matching and association verification to locate the fault type and generate a fault type list.

[0128] Step S4: Real-time acquisition of the frequency of the passage trigger signal reception, calculation of passage density and determination of the current passage flow level using the sliding window statistical method; when the fault type category or passage flow level is updated, the fault type list is weighted and sorted to determine the dominant fault type, the corresponding fault operation mode and flow operation mode are located respectively, the priority is determined according to the constraint strength of the operation mode parameters and the target operation mode is selected, the complete parameters of the target mode are extracted and sent to the status perception module, and the relevant parameters of the barrier drive command and reset command are dynamically fine-tuned.

[0129] Step S5: Receive the fault type list and current operating mode parameters; call the list of controlled fault types to match and identify key controlled fault types; generate targeted temporary control instructions based on the priority of fault impact; locate the associated components for each fault type through the fault component association mapping table; retrieve the cumulative runtime data of associated components; convert the total fault matching degree into a fault impact coefficient; call the corresponding time-series decay algorithm model to calculate the predicted remaining lifespan of associated components; classify the warning level according to the fault severity and remaining lifespan threshold; integrate the warning information and push fault details and maintenance suggestions to maintenance personnel through multiple channels; and provide temporary operation adjustment prompts to the target of passage.

[0130] Working principle and its effects:

[0131] This invention takes multi-module collaborative linkage as its core and builds a full-process intelligent management and control mechanism based on real-time feedback of operating status. Through closed-loop collaboration of status perception, analysis and diagnosis, feedback control and early warning interaction, it deeply integrates permission verification, operation adjustment, fault handling and early warning maintenance, realizing full-link intelligence of the barrier gate from trigger response to operation and maintenance support, fundamentally solving the pain points of insufficient precision in traditional equipment management, rigid operation and lagging fault handling.

[0132] After receiving the passage trigger signal, the status perception module collects multimodal recognition data and completes accurate permission verification through feature extraction and environmental adaptive weight adjustment. This effectively resists environmental interference such as light and weather, significantly reducing the risk of false interception or unauthorized passage and ensuring the accuracy of passage control. The analysis and diagnosis module accurately identifies operational anomalies and locates fault types through full-dimensional operation data collection, structured dataset construction, and multi-feature comparison. This breaks through the limitations of traditional single threshold judgment and enables early detection of potential faults. The feedback control module dynamically switches the operating mode based on the real-time passage flow level and the dominant fault type, enabling the equipment to adapt to different scenarios such as peak congestion, off-peak energy saving, and fault tolerance, balancing passage efficiency and equipment wear control. The early warning and interaction module generates targeted temporary control commands for control faults to suppress fault deterioration. At the same time, it uses a time-series decay algorithm combined with component operating data to predict the remaining lifespan and conduct graded early warnings, enabling timely fault handling and advance planning of maintenance plans, reducing passive downtime losses.

[0133] In summary, this invention constructs an intelligent control system that features precise permission verification, adaptive operation status, predictive fault handling, and proactive maintenance management. This system not only significantly improves the passage efficiency, operational stability, and control security of the barrier gate, but also reduces the impact of environmental interference, the risk of fault escalation, and maintenance costs. It effectively adapts to the diverse needs of parking lots, logistics parks, highways, and other scenarios, comprehensively enhancing the practicality and user experience of the equipment.

[0134] The above description is merely a preferred embodiment of the present invention. The scope of protection of the present invention is not limited to the above embodiments. All technical solutions falling within the scope of the present invention's concept are within the scope of protection of the present invention. It should be noted that for those skilled in the art, any improvements and modifications made without departing from the principles of the present invention should also be considered within the scope of protection of the present invention.

Claims

1. A barrier machine intelligent control system based on running state feedback, characterized in that, The state perception module, the analysis diagnosis module, the feedback control module and the early warning interaction module are included. The state perception module is configured to collect multi-modal recognition data when receiving a passage trigger signal, perform authentication feature extraction and permission rule base matching on the multi-modal recognition data, conduct dynamic permission verification, and determine whether to start the barrier machine operation. If so, the barrier machine control operation is performed based on the current operation mode of the barrier machine, barrier drive instructions and barrier reset instructions are generated, the barrier operation is controlled, and a feedback monitoring mechanism is started. The analysis diagnosis module is configured to collect barrier operation data when the feedback monitoring mechanism is started, construct an operation data set according to an operation period, identify an abnormal operation state in combination with a normal operation feature library of the current operation mode, and locate an abnormal operation feature. The abnormal operation feature is classified and matched with a fault rule library to locate a target fault rule and generate a fault type list. The feedback control module is configured to update the current barrier machine passage flow level according to the receiving frequency of the passage trigger signal, locate a dominant fault type from the fault type list, and dynamically switch the operation mode of the barrier machine in combination with the passage flow level. The early warning interaction module is configured to identify a management and control fault type in the fault type list, and generate temporary control instructions for each management and control fault type in turn.

2. The intelligent control system for barrier gate based on running state feedback according to claim 1, characterized in that, The residual life prediction value of the associated components of each fault type in the fault type list is calculated to generate fault anomaly early warning information and perform fault anomaly early warning. The step of determining whether to start the barrier machine operation includes: When receiving a passage trigger signal, multi-modal recognition data is collected and preprocessed, the multi-modal recognition data includes vehicle identity data, passage permission associated data and channel environment data; the passage trigger signal is an effective signal that can trigger the barrier machine to start the passage verification process; Based on the multi-modal recognition data after data preprocessing, authentication features are extracted and feature alignment processing is performed respectively, multi-dimensional feature vectors are constructed by splicing in the order of pre-set feature dimensions; The constructed multi-dimensional feature vectors are matched with the pre-set permission rule base, and the environmental impact level features corresponding to the channel environment data are called to dynamically adjust the weight proportion of each multi-dimensional feature vector dimension; 3. The intelligent control system for barrier gate based on running state feedback according to claim 2, characterized in that, The environmental impact level feature is a classification feature that reflects the degree of environmental interference with dynamic permission verification based on channel environment data and quantitative evaluation. The step of determining whether to start the barrier machine operation also includes: According to the weight proportion, the multi-dimensional feature vectors are compared with the pre-set constraint conditions in the permission rule base dimension by dimension, if all dimensions of the multi-dimensional feature vectors match the pre-set constraint conditions in the permission rule base, it is determined that the authentication is passed, and the barrier machine operation is started; If there are multi-dimensional feature vector dimensions that do not match the pre-set constraint conditions in the permission rule base, the sum of the weight proportions of the matched multi-dimensional feature vector dimensions is calculated as a feature matching index value. If the feature matching index value is greater than the preset feature matching threshold, it is determined that the authentication is passed, the barrier operation is started, and a review mark is made; otherwise, it is determined that the authentication fails, the barrier operation is prohibited, and an abnormal prompt warning is triggered synchronously.

4. The running state feedback based railing machine intelligent control system of claim 1, wherein, The step of performing the barrier control operation includes: When it is determined to start the barrier operation, a current operation mode parameter is received, and the operation mode parameter includes an action response priority, a power output constraint range, an action trajectory curvature interval, and an execution time reference; Based on the current operation mode parameter, a drive parameter set of the barrier drive instruction is generated, and then a barrier drive instruction adapted to the current operation mode is generated; The barrier drive instruction includes an action trajectory parameter and a power output parameter, the action trajectory parameter includes an upper limit of a lifting angle, a rate change gradient, and a trajectory smoothing coefficient, and the power output parameter includes a motor voltage adjustment range, a torque real-time compensation value, and a start-stop acceleration threshold; The barrier drive instruction is sent to the barrier execution device to drive the barrier execution device to operate according to the barrier drive instruction, and a feedback monitoring mechanism is started at the same time; After the barrier completes the lifting action, channel state data is collected, a channel state time series data set is constructed, and a passing feature is extracted, which is used to determine whether the target passing object has left the channel monitoring area where the current barrier machine is located and whether there is an obstacle in the channel monitoring area where the current barrier machine is located; If the target passing object has left the channel monitoring area where the current barrier machine is located and there is no obstacle in the channel monitoring area, a barrier reset operation is started, a barrier reset instruction is generated, the barrier reset operation is driven to operate, and a feedback monitoring mechanism is started.

5. The running state feedback based railing machine intelligent control system of claim 1, wherein, The step of identifying the abnormal running state and locating the abnormal running feature includes: When the feedback monitoring mechanism is started, barrier operation data is collected at a preset sampling frequency, and the barrier operation data includes power system running data, mechanical action feedback data, sensor sensing data, and instruction execution response data; After the collected barrier operation data is classified and preprocessed, a structured running data set is constructed according to a single operation period; multi-dimensional running features are extracted from the running data set; A preset normal running feature library is called, and the normal running feature library includes running feature standard intervals of each dimension classified according to the running mode and feature correlation rules; The extracted multi-dimensional running features are compared with the running feature standard intervals of the corresponding running mode in the normal running feature library dimension by dimension, the deviation degree of the real-time running features and the center value of the running feature standard interval is calculated, and the matching degree of each real-time running feature combination and the corresponding feature correlation rule is calculated; The deviation judgment threshold and the matching degree judgment threshold are set, if the deviation degree of a single-dimensional running feature exceeds the deviation judgment threshold, or the matching degree of a multi-dimensional running feature combination is lower than the matching degree judgment threshold, it is determined that the current barrier machine has a suspected running abnormal state, and a suspected abnormal running feature is located; If the current feedback monitoring mechanism detects a suspected abnormal running feature, the frequency of the operation period in which the suspected abnormal running feature appears in a preset evaluation period is called from the running data set, and if it is greater than a preset frequency threshold, it is determined that the current barrier machine has a running abnormal state, and the suspected abnormal running feature is marked as an abnormal running feature.

6. The running state feedback based railing machine intelligent control system of claim 1, wherein, The step of generating the fault type list according to the positioning target fault rule comprises: When the abnormal operation state is identified, based on the operation data set, the abnormal operation features marked in the evaluation period and the abnormal operation feature values of the corresponding operation period are obtained, and numerical interval scaling is performed to construct a standardized abnormal feature vector; According to the data sub-module to which the abnormal operation feature belongs, the abnormal feature vector is classified, and an abnormal feature vector subset is constructed; the data sub-module comprises a power system data sub-module, a mechanical action data sub-module, a sensor data sub-module, and an instruction response data sub-module; For each abnormal feature vector subset, the fault rules of the preset fault rule library are screened according to the data sub-module label to which the abnormal feature vector subset belongs, and a fault rule subset corresponding to the data sub-module is constructed; the preset fault rule library stores the fault rules according to the data sub-module, and each fault rule comprises a fault type, a typical abnormal feature combination, and a parameter determination threshold in the corresponding data sub-module; For each abnormal feature vector subset and the corresponding fault rule subset, feature matching is performed, the total matching degree of each abnormal feature vector subset and the corresponding fault rule subset is calculated, and the fault rules with a total matching degree higher than a preset matching threshold are marked as candidate fault rules; For each candidate fault rule, the associated verification parameters corresponding to the candidate fault rule are retrieved from the operation data set; the associated verification parameters are compared with the preset normal operation range to verify the target fault rule; and the target fault rules are sorted in descending order of the total matching degree, and the fault types corresponding to the target fault rules are sequentially given to generate a fault type list.

7. The running state feedback based railing machine intelligent control system of claim 6, wherein, The step of calculating the total matching degree of each abnormal feature vector subset and the corresponding fault rule subset comprises: The number of feature identifiers matched with the typical abnormal feature combination of each fault rule in the fault rule subset in the abnormal feature vector subset is counted, and the proportion of the number of feature identifiers matched with the typical abnormal feature combination of each fault rule in the fault rule subset in the abnormal feature vector subset is counted. For each matched feature identifier, the numerical deviation degree of the abnormal operation feature in the abnormal feature vector is compared with the parameter determination threshold preset by the fault rule, and the parameter deviation fit degree score is generated according to the deviation amplitude ratio of the numerical deviation degree to the parameter determination threshold range; the average value of the parameter deviation fit degree scores of all matched features is taken as the comprehensive parameter deviation fit degree; The weights of the feature identifier matching degree and the comprehensive parameter deviation fit degree are set, and the feature identifier matching degree and the comprehensive parameter deviation fit degree are weighted and summed according to the corresponding weights to obtain the total matching degree.

8. The running state feedback based railing machine intelligent control system of claim 1, wherein, The step of dynamically switching the operation mode of the barrier machine comprises: The receiving frequency of the barrier machine passage trigger signal is obtained in real time, the passage density of the barrier machine per unit time is calculated by using a sliding window statistical method, the passage density is compared with a preset flow level threshold, and the current passage flow level is obtained; When the barrier machine fault type category is updated or the passage flow level is updated, a barrier machine operation mode switching operation is triggered: The fault type list is received, the influence degree priority of the fault type is combined, the total matching degree of each fault type is calculated, the comprehensive ranking score of each fault type is calculated, the fault type list is sorted, and the dominant fault type is located; According to the preset fault mode mapping rule, the current traffic level, and the preset traffic mode mapping rule, the corresponding fault operation mode and traffic operation mode are located; and based on the preset operation mode priority, the target operation mode is selected from the fault operation mode and the traffic operation mode.

9. The intelligent barrier machine control system based on operating condition feedback of claim 1, wherein, The step of generating fault abnormality early warning information and performing fault abnormality early warning includes: Receiving a fault type list, calling a preset control fault type list, and matching each fault type in the fault type list one by one to identify the current control fault type of the barrier machine; According to the influence degree priority of the control fault type, the temporary control instruction of each control fault type is generated in turn combined with the current operation mode parameter; Call the preset fault component association mapping table to locate the associated components of each fault type, and call the cumulative running time data of each associated component to generate the running time sequence of each associated component; The total matching degree of each fault type is converted into a fault influence coefficient, which is used to quantify the additional attenuation of the current fault type on the associated component life; For each associated component, call the adaptive timing attenuation algorithm model, take the design life benchmark value of the associated component as the initial value, calculate the normal attenuation under the fault-free state based on the running time sequence, superimpose the additional attenuation of the current fault type, obtain the cumulative loss amount of the associated component, and then obtain the residual life prediction value of the associated component of each fault type; According to the influence degree priority of the control fault type and the residual life prediction value of the associated component of each fault type, the early warning level is divided, and the fault abnormality early warning information is integrated to perform fault abnormality early warning.

10. A barrier machine intelligent control method based on operation state feedback, implemented based on the barrier machine intelligent control system based on operation state feedback in any one of claims 1-9, characterized in that, It includes: Step S1: When receiving a traffic trigger signal, collect multi-modal recognition data, perform authentication feature extraction and permission rule library matching on the multi-modal recognition data, perform dynamic permission verification, and determine whether to start the barrier machine operation; Step S2: If the barrier machine operation is started, execute the barrier machine control operation based on the current operation mode of the barrier machine, generate the barrier drive instruction and the barrier reset instruction, control the barrier operation, and start the feedback monitoring mechanism; Step S3: When the feedback monitoring mechanism is started, collect barrier operation data, construct a running data set according to the operation period, identify the abnormal running state combined with the normal running feature library of the current operation mode, and locate the abnormal running feature; And the abnormal running feature is classified and matched with the fault rule library to locate the target fault rule and generate a fault type list; Step S4: According to the receiving frequency of the traffic trigger signal, update the current barrier machine traffic level, and locate the dominant fault type from the fault type list to dynamically switch the operation mode of the barrier machine combined with the traffic level; Step S5: Identify the control fault type in the fault type list, and generate the temporary control instruction of each control fault type in turn; The remaining life prediction values of the associated components of each fault type in the fault type list are calculated to generate fault anomaly early warning information and perform fault anomaly early warning.

Images